Flow rate control device and flow rate control method

Abstract

To provide a flow rate controller with which responsiveness to falling transitions has been improved, while the occurrence of undershoots is prevented.SOLUTION: A flow rate controller 8 comprises a diaphragm part 2, a control valve 6 on the diaphragm part upstream side, an upstream pressure sensor 3 between the diaphragm part and the control valve, and a control mechanism 7 connected to the control valve and the upstream pressure sensor, and is constituted so as to control the control valve for controlling a flow rate on the basis of the output of the upstream pressure sensor. When performing operation to close the control value in order to lower the flow rate of a gas flowing to the downstream side of the diaphragm part from an initial value A% to a final target value B%, the control valve is controlled on the basis of an internal flow rate signal that updates a control target value at a prescribed interval, with the internal flow rate signal generated on the basis of a control formula below that includes a time constant τ2: control target value=previous target value+(final target value-previous target value) / (τ2+1), where the time constant τ2 is smaller than a time constant τ1 at the time of lowering from the initial value A% to a final target value 0%.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a flow rate control device and a flow rate control method, and more particularly to a flow rate control device and a flow rate control method that can improve flow rate reduction responsiveness.

[0002] In semiconductor manufacturing equipment and chemical plants, various flow meters and flow controllers are used to control the flow rates of material gases and etching gases. Among these, pressure-type flow controllers are widely used because they can control the mass flow rate of various fluids with high precision using a relatively simple mechanism that combines a control valve and a restrictor (e.g., an orifice plate or a critical flow nozzle). Unlike thermal-type flow controllers, pressure-type flow controllers have excellent flow control characteristics, such as stable flow rate control even when the primary supply pressure upstream of the control valve fluctuates significantly (e.g., Patent Document 1).

[0003] Some pressure-type flow control devices adjust the flow rate by controlling the fluid pressure upstream of the throttle (hereinafter sometimes referred to as the upstream pressure P1). The upstream pressure P1 is usually controlled by adjusting the aperture of a control valve located upstream of the throttle. When the critical expansion condition (upstream pressure P1 / downstream pressure P2 ≧ approximately 2: in the case of argon gas) is met, the velocity of the gas flowing through the throttle is maintained at the speed of sound, and the mass flow rate of the gas flowing downstream of the throttle is proportional to the upstream pressure P1, regardless of the magnitude of the downstream pressure P2 downstream of the throttle. Therefore, the flow rate can be controlled by controlling the upstream pressure P1 using a control valve. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] International Publication No. 2015 / 083343 [Patent Document 2] International Publication No. 2018 / 180745 Summary of the Invention [Problem to be solved by the invention]

[0005] However, in a pressure-type flow control device, a throttle is provided downstream of the control valve, so even after the control valve is closed, residual gas between the control valve and the throttle flows downstream through the throttle. For this reason, even if the control valve is quickly closed, the flow rate does not decrease rapidly, and it tends to take a relatively long time for the flow rate to decrease. In order to improve the responsiveness of the flow rate decrease, it is possible to design the flow path volume between the control valve and the throttle as small as possible, but this requires the connection of a pressure sensor, and there is a limit to how small the volume can be made.

[0006] On the other hand, in a pressure-type flow control device, when a step-like target flow rate signal that rapidly reduces the flow rate is given, the control valve may temporarily close excessively, causing undershoot. This is because the opening of the control valve is controlled by feedback control based on the upstream pressure P1, so the control valve may remain fully closed until the upstream pressure P1 reaches the target pressure, which may cause a delay in the subsequent opening control.

[0007] To address this problem, Patent Document 2 by the present applicant discloses a flow control device that uses an internal flow control signal that exponentially attenuates a target flow rate and closes a control valve through first-order delay control, thereby preventing undershoot and reducing the flow rate without impairing responsiveness as much as possible.The flow control device described in Patent Document 2 prevents the control valve from being fully closed and prevents undershoot from occurring by not making the value of the time constant τ of the first-order delay control, which determines the attenuation rate of the target flow rate, too small.

[0008] However, in the flow control device described in Patent Document 2, the internal flow control signal was generated in the same manner regardless of the target flow rate after the shutdown, so even though undershoot could be prevented, there was still room for improvement in responsiveness.

[0009] The present invention has been made in consideration of the above-mentioned problems, and its main object is to provide a flow control device and a flow control method that can improve the responsiveness of the flow rate decrease while preventing the occurrence of undershoot. [Means for solving the problem]

[0010] A flow control device according to an embodiment of the present invention comprises a throttling section, a control valve provided upstream of the throttling section, an upstream pressure sensor that detects the pressure between the throttling section and the control valve, and a control mechanism connected to the control valve and the upstream pressure sensor, and is configured to control flow by controlling the control valve based on the output of the upstream pressure sensor, and when closing the control valve to reduce the flow rate of gas flowing downstream of the throttling section from an initial value A% to a final target value B%, the opening of the control valve is controlled based on an internal flow signal that updates the control target value at predetermined intervals, and the internal flow signal is generated based on the following control equation including a time constant τ2: Control target value = previous target value + (final target value - previous target value) / (τ2 + 1), where the time constant τ2 is smaller than the time constant τ1 when reducing from the initial value A% to the final target value 0%.

[0011] In one embodiment, the time constant τ2 is a fixed value over the period in which the gas flow rate is reduced from an initial value A% to a final target value B%, and is determined by τ2=τ1×(AB) / AB / A.

[0012] In one embodiment, the time constant τ2 is updated every time the control target value is updated at the predetermined interval during the period in which the gas flow rate is reduced from an initial value A % to a final target value B %.

[0013] A flow rate control method according to an embodiment of the present invention is executed in a flow rate control device including a throttle section, a control valve provided upstream of the throttle section, an upstream pressure sensor that detects the pressure between the throttle section and the control valve, and a control mechanism connected to the control valve and the upstream pressure sensor, and includes the steps of receiving a signal to reduce the flow rate of gas flowing downstream of the throttle section from an initial value A% to a final target value B%, generating an internal flow rate signal for controlling the opening of the control valve when the signal to reduce the flow rate is received, and adjusting the opening of the control valve in accordance with the internal flow rate signal. and a step of feedback-controlling the opening of the control valve based on the output of an upstream pressure sensor, wherein the internal flow rate signal is generated based on the following control equation including a time constant τ2 as a signal for updating the control target value at predetermined intervals: Control target value = previous target value + (final target value - previous target value) / (τ2 + 1), where the time constant τ2 is smaller than the time constant τ1 when decreasing from the initial value A% to the final target value 0%, but is set so that the curve defined by the control equation using the time constant τ2 exceeds the curve defined by the control equation using the time constant τ1 at any time. [Effects of the Invention]

[0014] According to the flow rate control device and flow rate control method of the embodiment of the present invention, it is possible to improve the responsiveness of the flow rate fall while preventing the occurrence of undershoot. [Brief explanation of the drawings]

[0015] [Figure 1] 1 is a schematic diagram showing a flow control device according to an embodiment of the present invention; [Figure 2] 10 is a graph showing a first-order lag flow control signal generated when the flow rate is reduced. [Figure 3] 10 is a graph showing an internal flow control signal of a comparative example. [Figure 4] 10 is a graph showing an internal flow control signal according to an embodiment. [Figure 5] 4 is a flowchart showing a flow rate control operation of the embodiment. [Figure 6] 10 is a graph illustrating an internal flow control signal according to another embodiment. [Figure 7] 10 is a flowchart showing a flow rate control operation according to another embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0016] Hereinafter, embodiments of the present invention will be described with reference to the drawings, but the present invention is not limited to the following embodiments.

[0017] 1 shows the configuration of a flow rate control device 8 according to an embodiment of the present invention. The flow rate control device 8 is a pressure-type flow rate control device that includes a throttle section 2 (e.g., an orifice plate) interposed in a flow path (gas supply path) 1 through which gas G passes, a control valve 6 provided upstream of the throttle section 2, an upstream pressure sensor 3 and a temperature sensor 5 provided between the control valve 6 and the throttle section 2, and a downstream pressure sensor 4 provided downstream of the throttle section 2. The upstream pressure sensor 3 measures an upstream pressure P1, which is the fluid pressure between the control valve 6 and the throttle section 2, and the downstream pressure sensor 4 measures a downstream pressure P2, which is the fluid pressure between the throttle section 2 and a shut-off valve 9.

[0018] The upstream side of the flow rate control device 8 is connected to a gas supply source (not shown) for a material gas, etching gas, carrier gas, or the like, and the downstream side is connected to a process chamber 10 of a semiconductor manufacturing device via a shutoff valve 9. A vacuum pump 11 is connected to the process chamber 10, and the inside of the process chamber 10 and the flow paths connected thereto can be evacuated when gas G is supplied. In the embodiment shown in FIG. 1, the shutoff valve 9 is disposed outside the flow rate control device 8, but the shutoff valve 9 may also be built into the flow rate control device 8. As the shutoff valve 9, for example, an on-off valve such as an AOV (air-operated valve) or a solenoid valve is suitably used.

[0019] In the flow control device 8, the upstream pressure sensor 3 and the downstream pressure sensor 4 are, for example, semiconductor piezo-resistance diffusion pressure sensors or capacitance manometers, and the temperature sensor 5 is, for example, a resistance temperature detector or thermistor. The control valve 6 is, for example, a piezoelectric element-driven valve (hereinafter sometimes referred to as a piezo valve) in which a metal diaphragm valve element is opened and closed by a piezo actuator. A piezo valve is a valve (proportional valve) that can be opened to any opening degree by adjusting the drive voltage applied to the piezo element. The throttle section 2 is, for example, an orifice plate or a critical nozzle. The opening diameter of the throttle section 2 is set to, for example, 10 to 2000 μm.

[0020] The flow control device 8 also includes a control mechanism (or control circuit) 7 that controls the opening and closing operation of the control valve 6 based on the outputs of the upstream pressure sensor 3 and the downstream pressure sensor 4. The control mechanism 7 compares a set flow rate determined by an external command signal from the external control device 12 with a calculated flow rate calculated from the output of the upstream pressure sensor 3, and performs feedback control on the control valve 6 so that the calculated flow rate approaches the set flow rate, i.e., so that the difference between the calculated flow rate and the set flow rate approaches zero.

[0021] The control mechanism 7 is typically built into the flow control device 8, but may also be provided externally to the flow control device 8. The control mechanism 7 typically includes a CPU, a memory (storage device) M such as ROM or RAM, an A / D converter, and the like, all mounted on a circuit board, and may include a computer program configured to execute the flow control operation described below. The control mechanism 7 is realized by a combination of hardware and software. The control mechanism 7 may be provided with an interface for exchanging information with an external device such as a computer, and programs and data can be written to the ROM from the outside. The components of the control mechanism 7 do not all need to be provided integrally; some components, such as the CPU, may be located in different locations. These components may communicate with each other via wire or wirelessly.

[0022] Unlike the illustrated embodiment, the flow rate control device 8 may not include the downstream pressure sensor 4. In this case, the control mechanism 7 is configured to calculate the flow rate based on the output of the upstream pressure sensor 3. The control mechanism 7 may also be configured to correct the calculated flow rate based on the fluid temperature detected by the temperature sensor 5. Furthermore, the flow rate control device 8 may include an inlet-side pressure sensor (not shown) upstream of the control valve 6 for measuring the gas supply pressure. The inlet-side pressure sensor can measure the pressure of the gas supplied from a connected gas supply device (e.g., a raw material vaporizer) and is used to control the primary-side gas supply pressure, etc.

[0023] In the pressure-type flow control device 8 described above, when the critical expansion condition P1 / P2 ≧ approximately 2 (in the case of argon gas) is satisfied, the mass flow rate of the gas flowing downstream of the throttle section 2 is determined by the upstream pressure P1, not by the downstream pressure P2. When the critical expansion condition is satisfied, the flow rate Q downstream of the throttle section 2 is given by Q = K1 · P1 (K1 is a constant that depends on the type of fluid and the fluid temperature), and the flow rate Q is proportional to the upstream pressure P1 measured by the upstream pressure sensor 3. Furthermore, when the downstream pressure sensor 4 is provided, the difference between the upstream pressure P1 and the downstream pressure P2 is small, and the flow rate can be calculated even when the above critical expansion condition is not satisfied, and Q = K2 · P2 is calculated based on the measured upstream pressure P1 and downstream pressure P2. m (P1-P2) n (where K2 is a constant that depends on the type of fluid and the fluid temperature, and m and n are exponents derived based on the actual flow rate) The flow rate Q can be calculated from

[0024] To control the flow rate, the set flow rate set in the external control device 12 is sent from the external control device 12 to the control mechanism 7. The control mechanism 7 calculates the flow rate using the above flow rate calculation formula: Q=K1·P1 or Q=K2·P2 under the critical expansion condition or the non-critical expansion condition. m (P1-P2) nThe control valve 6 is feedback-controlled based on the output of the upstream pressure sensor 3 so that the flow rate of the fluid passing through the throttle section 2 approaches the set flow rate (i.e., so that the difference between the calculated flow rate and the set flow rate approaches 0). The calculated flow rate may be output to an external control device 12 and displayed as a flow rate output value.

[0025] The operation of the flow control device 8 when the flow rate is reduced will be described below. In this specification, all flow rate values ​​such as the set flow rate and target flow rate or upstream pressure values ​​may be expressed as a ratio where a predetermined flow rate value (typically a rated flow rate value) or a corresponding upstream pressure value is set to 100%.

[0026] FIG. 2 shows the flow rate target signal A2 generated inside the flow control device 8 in response to a flow rate setting signal A1 that reduces the flow rate in a stepwise manner received from the outside. As shown in FIG. 2, the flow rate target signal A2 is generated so as to decay exponentially, causing the control valve 6 to close under first-order lag control. Even when the flow rate is reduced, the drive voltage of the control valve 6 fluctuates as needed so that the difference between the calculated flow rate determined from the output of the upstream pressure sensor 3 and the set flow rate target value A2 becomes zero. Here, the transfer function G(s) of the first-order lag system is expressed as G(s) = K / (τs+1), where K is the gain and τ is the time constant. The rising step response of the first-order lag system is expressed as the exponential function y(t) = K(1-exp(-t / τ)), which can be derived by performing an inverse Laplace transform on the above transfer function.

[0027] Furthermore, in this embodiment, the flow rate target signal A2 is generated as a signal that does not fall below a reference flow rate decrease characteristic equivalent to the reference pressure decrease characteristic, i.e., a signal that is slower than the reference flow rate decrease characteristic. Here, the reference pressure decrease characteristic refers to the pressure decrease characteristic when the residual gas between the control valve 6 and the throttle unit 2 flows downstream through the throttle unit 2 after the control valve 6 is instantaneously closed, and the reference flow rate decrease characteristic is similar, except that it is converted into a flow rate. The reference pressure decrease characteristic can be considered to represent the fastest flow rate decrease characteristic in the flow control device 8.

[0028] If the flow rate target signal A2 falls below the reference flow rate reduction characteristic, the control valve 6 will remain closed, effectively resulting in an uncontrolled state. This can cause undershoot. Therefore, by generating the flow rate target signal A2 so that it does not fall below the reference flow rate reduction characteristic, the control valve 6 can be controlled to close in accordance with the flow rate target signal A2 without falling into an uncontrolled state.

[0029] The reference pressure drop characteristic is typically expressed as P(t) = P0·exp(-t / τ), which represents exponential decay. The reference flow rate decline characteristic is expressed as Q(t) = Q0·exp(-t / τ). Here, P(t) and Q(t) are functions of the upstream pressure P1 and flow rate Q with respect to time t, P0 and Q0 are the initial upstream pressure and initial flow rate, respectively, and τ is a time constant (which is essentially the same value in both equations) indicating the rate of decay. The smaller the time constant τ, the faster the decay occurs, and the larger the time constant τ, the slower the decay occurs. More specifically, in exponential decay, the time constant τ corresponds to the time required for the output to decline to approximately 36.8% (the reciprocal of Napier's number e) from an initial value of 100%. In the flow control device 8, the time constant τ of the reference pressure drop characteristic or reference flow rate decline characteristic can take various values ​​depending, for example, on the diameter of the throttle section 2 and the type of gas.

[0030] Therefore, by performing first-order lag control in which the time constant τ of the reference flow rate reduction characteristic in the flow control device 8 is determined in advance by measurement and the time constant τ of the flow rate target signal A2 is set to be larger than the time constant τ of the reference flow rate reduction characteristic, it is possible to prevent the control valve 6 from going into an uncontrolled state and to perform flow rate control in accordance with the flow rate target signal A2.

[0031] Furthermore, by setting the time constant τ of the flow rate target signal A2 to a value larger than the time constant τ of the reference flow rate reduction characteristic by providing a larger margin, it is possible to absorb the difference in characteristics between the devices of the flow rate control device, and to eliminate the variation in responsiveness between the flow rate control devices and unify them.

[0032] Furthermore, when the target value is set and updated at a predetermined control interval, the internal flow control signal for performing the above-mentioned first-order lag control can be generated according to the following control formula. Control target value = previous target value + (final target value - previous target value) / (time constant + 1)

[0033] The above control equation is A n is the nth control target value (current value), and A n-1 The target value for the n-1th time, i.e., A n is the previous value of , a constant final target value which is an external command value is B, and τ is a time constant (parameter), it is expressed by the following equation. A n =A n-1 +(BA n-1 ) / (τ+1)

[0034] In the above formula, the final target value B is a constant value, and the previous value A n-1 The change in (i.e., (BA n-1 The absolute value of (τ) / (τ+1) decreases as time passes, corresponding to the previous value decreasing, which corresponds to first-order lag control. In first-order lag control, the amount of change per unit time is greatest at the start of the flow rate reduction, and as time passes, the amount of change decreases and the curve gradually becomes gentler, until after a sufficient amount of time has passed, the curve asymptotically approaches the final target value B. Furthermore, the time constant τ, which can be set as a parameter, becomes larger, resulting in a more gradual falling curve, and the smaller this is, the more steeply the curve falls.

[0035] Figure 3 is a graph showing a comparative example of the internal flow control signal of the first-order lag system according to the above formula. Here, graphs C0, C1, C2, C3, and C4 are shown when transitioning from a flow rate setting of 100% before attenuation to flow rates of 0%, 10%, 20%, 40%, and 80%, which are the final target value B.

[0036] 3, the time constant τ1 used in the above equation when the flow rate is reduced from 100% to 0%, i.e., the time constant τ1 in graph C0, is used regardless of the magnitude of the final target value. Here, as described above, the time constant τ1 in graph C0 is set to a value slightly larger than the time constant of the reference flow rate reduction characteristic (for example, 105 to 120% of the reference) in order to prevent undershoot.

[0037] For example, when performing first-order delay control from an initial value of 100% to a final target value of 0% at 0.5 ms intervals, if the initial control target value A1 is given as A1 = 100 + (0 - 100) / (τ + 1) and the time constant τ is set to 99, the control target value A2 after 0.5 ms will be A2 = 99 + (0 - 99) / (τ + 1). If this is repeated up to the final target value of 0%, a graph showing exponential decay like graph C0 in Figure 3 will be obtained.

[0038] Furthermore, even when the final target values ​​are different, if the same common time constant τ1 (99 in the above example) is applied, as can be seen from the above equation, attenuation occurs with an amount of change proportional to the setting difference between the initial value and each final target value, so each graph C0 to C4 will take on a value given at the same ratio to the setting difference at the same time.

[0039] For example, if at a certain time t the value of graph C0 for a final target value of 0% is 50% (half), then at the same time t the value of graph C1 for a final target value of 10% (i.e., a setting difference of 90%) will be 100-90 / 2=55%, the value of graph C2 for a final target value of 20% (i.e., a setting difference of 80%) will be 100-80 / 2=60%, the value of graph C3 for a final target value of 40% (i.e., a setting difference of 60%) will be 100-60 / 2=70%, and the value of graph C4 for a final target value of 80% (i.e., a setting difference of 20%) will be 100-20 / 2=90%, and graphs C1 to C4 will form curves that compress graph C0 along the vertical axis according to the setting difference.

[0040] As a result, although there is an advantage that a roughly constant fall time can be obtained regardless of the final target value, as is clear from graph C4, when the final target value is large, the decay occurs at a rate slower than the limit rate considering the reference pressure drop characteristic, and the fall time becomes longer than necessary. In reality, when the final target value is not 0%, processing in the process chamber is often ongoing, so it is often necessary to fall the flow rate to the desired value as quickly as possible, but this cannot be achieved if the same common time constant τ1 is used.

[0041] Therefore, in the embodiment specifically described below, instead of using a fixed first-order lag time constant, the first-order lag time constant is changed depending on the final target value, thereby controlling the flow rate so that each final target value is reached as quickly as possible while preventing undershoot.

[0042] Fig. 4 is a graph of an embodiment showing an internal flow control signal of a first-order lag system, and similarly to Fig. 3, solid lines indicate graphs E0, E1, E2, E3, and E4 when transitioning from a flow rate setting of 100% before attenuation to flow rate settings of 0%, 10%, 20%, 40%, and 80%, which are the final target value B. Note that graphs C0 to C4 of the comparative example shown in Fig. 3 are indicated by dashed lines in this figure.

[0043] However, unlike the comparative example, graphs E0, E1, E2, E3, and E4 of the embodiment shown in Fig. 4 use a variable first-order lag time constant, and a different time constant τ2 is used for each final target value B. More specifically, when the final target value B is 0%, the time constant τ1 in graph C0 shown in Fig. 3 is used as is as the time constant τ2, while as the final target value B increases to 10%, 20%, 40%, and 80%, individual time constants τ2 having smaller values ​​are used.

[0044] The time constant τ2 appropriately set for each final target value B can be found, for example, as follows. First, in either case, to prevent undershoot, it is required that the command value does not fall below the falling waveform of the limit speed corresponding to the reference drop characteristic. Furthermore, when changing the set flow rate from the initial value A% (100% in the above example) to the final target value B%, it is required that the falling curve Y1(t) from the set A% to 0% with the time constant τ1 and the falling curve Y2(t) from the set A% to B% with the time constant τ2 do not fall below the former Y1(t) at any time t.

[0045] τ2, which satisfies the above conditions, can be defined, for example, by the following equation (1) using the time constant τ1, initial value A (%), and final target value B (%) when attenuating from the set A% to 0%. τ2=τ1×(AB) / AB / A (1)

[0046] The derivation process of equation (1) will be explained below. Here, the time constant τ2 is set to a different value depending on the final target value B, while a constant value is used regardless of time in the damping process having each final target value B. Also, here, a case where an initial value A0 immediately before the start of damping is used will be explained, but the same explanation below also applies when performing control toward the target value B from an arbitrary starting value A0.

[0047] On the curve Y2(t) where the final target value is other than 0%, the control target value A moves from the initial value A0 to the target value B. n is expressed by the following formula: A n =A n-1 -(A n-1 -B) / (τ2+1) (2a) On the other hand, in the curve Y1(t) where the final target value is 0%, the control target value A' approaches the target value 0 from the initial value A0. n is expressed by the following formula: A' n =A' n-1 -A' n-1 / (τ1+1) (2b)

[0048] Here, let us consider the magnitude of the control command value when the first control period has elapsed (first cycle). The value A1 of the first cycle corresponding to the curve Y2(t) is A1 = A0 - (A0 - B) / (τ2 + 1), and the value A'1 of the first cycle corresponding to the curve Y1(t) is B = 0. Substituting these, A'1=A0-A0 / (τ1+1).

[0049] In order for Y2(t) to not fall below Y1(t), A1 ≥ A'1 must be satisfied. Therefore, it is necessary to satisfy A0-(A0-B) / (τ2+1)≧A0-A0 / (τ1+1). By rearranging this equation, we obtain the following equation (3) from τ2+1≧(A0-B) / A0×(τ1+1). τ2≧τ1(A0-B) / A0-B / A0...(3)

[0050] Therefore, in the first cycle, the minimum value of τ2 that satisfies the condition is expressed as τ2 = τ1(A0 - B) / A0 - B / A0, which corresponds to the above equation (1). Note that when τ2 = τ1(A0 - B) / A0 - B / A0, A1 = A'1.

[0051] When τ2 that satisfies equation (3) is used, in exponential decay where the decay rate is determined by the time constant τ, Y2(t) ≥ Y1(t) holds even at any time t from the second cycle onwards, and in particular, when B ≠ 0, Y2(t) > Y1(t) holds (when B = 0, Y2(t) = Y1(t)). This is because, as can be seen from equations (2a) and (2b) above and the graph in Figure 4, in the exponential decay of this example, the difference in the amount of change per cycle increases over time depending on the difference in the time constant τ and the setting difference, and Y2(t) and Y1(t) have a relationship where they gradually diverge greatly.

[0052] Therefore, if the minimum value of τ2 obtained from equation (3) is used, Y2(t) will not fall below Y1(t) in the process of transitioning from A to B. In other words, if the time constant τ2 = τ1 × (AB) / AB / A defined in equation (1) is used, the curve of the time constant τ2 when decreasing from the set A% to B% will not fall below the curve of the time constant τ1 from the set A% to 0%, and the final target value B can be reached relatively quickly without undershooting.

[0053] Furthermore, the time constant τ2 defined by equation (1) takes a different value depending on the magnitude of the final target value B, and typically, the larger the final target value B, the smaller the value of τ2. For example, if the time constant τ1 is 100 when B=0, the time constant τ2 when the final target value is 10% is τ2=100×(100-10) / 100-10 / 100=89.9, whereas the time constant τ2 when the final target value is 40% is τ2=100×(100-40) / 100-40 / 100=59.6, and the time constant τ2 when the final target value is 80% is τ2=100×(100-80) / 100-80 / 100=19.2. This means that the closer the final target value is to 0%, the more it is necessary to use a time constant τ2 that is closer to the time constant τ1, and the higher the final target value is, the more room there is for shortening the fall time by using a smaller time constant τ2.

[0054] In this way, in the above embodiment in which the time constant τ2 is determined by equation (1), the time constant τ21 used when the final target value B% is the first final target value B1% (e.g., 80%) is set to be smaller than the time constant τ22 used when the final target value B% is the smaller second final target value B2% (e.g., 20%). In other words, the time constant τ2 that enables the final target value B to be reached as quickly as possible while preventing undershoot is set individually for each final target value B.

[0055] A specific example of a flow rate control method will be described below with reference to the flowchart shown in Fig. 5. The flow rate control operation by first-order delay control for performing step-down is performed by adjusting the current control value A n The control is initiated when a final target value B smaller than the reference value B is given, and is performed by the control loop shown in FIG.

[0056] First, as shown in step S1 of FIG. 5, it is determined whether the number of cycles n since the start of first-order lag control is 1, i.e., whether this is the first cycle of step-down flow rate control. If n=1, then as shown in step S2, a time constant τ2 based on the final target value B is calculated. The time constant τ2 is determined, for example, based on the above equation (1) using the initial value A, the final target value B, and a time constant τ1 that has been measured in advance and stored in memory and that can prevent undershoot when changing from the set value A% to 0%. Note that if it is determined in step S1 that n=1, and this is not a cycle in which first-order lag control is being executed using the already determined time constant τ2, then step S2 for determining the time constant τ2 is skipped, and the process proceeds to step S3 for determining the next control target value.

[0057] In step S3, the control target value is updated by first-order delay control. Specifically, by substituting τ=τ2, as shown in the above equation (2a), A n =A n-1 -(A n-1 -B) / (τ2+1). Also, as shown in step S4, n is incremented and the determined control target value is designated as the next control target value.

[0058] Thereafter, as shown in step S5, the control target value A obtained in step S3 is n The feedback control of the control valve 6 is performed based on the output of the upstream pressure sensor. n This is done by adjusting the opening to suit the

[0059] By repeating this control loop, the time constant τ2, which is appropriately set for the final target value B, is used to prevent the control valve 6 from going out of control and to prevent undershoot, while allowing the flow rate to step down to the final target value B more quickly than when time constant τ1 is used.

[0060] A flow rate control device and a flow rate control method according to another embodiment will now be described.

[0061] In the above embodiment, first-order delay control was performed using a time constant τ2 that was set to a different value depending on the final target value B, but a fixed value was used for the time constant τ2 during each step-down control. In contrast, in another embodiment described below, by performing control to recalculate the minimum value of the applicable time constant τ2 even during each step-down control, it is possible to achieve control that draws a curve that is closer to the curve of the time constant τ1 from the set A% to 0%.

[0062] 6 is a graph showing the internal flow control signal of the first-order lag system in another embodiment, and similar to Figures 3 and 4, graphs E0, E11, E12, E13 and E14 are shown when transitioning from a flow rate setting of 100% before attenuation to flow rate settings of 0%, 10%, 20%, 40% and 80%, which are the final target value B. Note that graph E0 showing the transition to 0% is the same as graph C0 in the comparative example and graph E0 in the embodiment described above.

[0063] As shown in Fig. 6, in this embodiment, by repeatedly recalculating the time constant τ2 during the step-down, all of the graphs E11 to E14 show curves that are almost the same as the graph E0 during the attenuation period. However, when the control target value reaches the final target value B, as can be seen from the above formula (2a), the control change amount becomes 0, and thereafter, the control target value A, which is the same as the final target value B, n is given continuously.

[0064] The real-time update of the time constant τ2 is, for example, in the above equation (1), instead of the fixed initial value A, the current control target value An By using the current time constant τ2, which is updated in real time based on the following equation (4): n can be obtained. τ2 n =τ1×(A n-1 -B) / A n-1 -B / A n-1 ···(4) In this way, by updating the time constant τ2 in real time, the fall time for each final target value B can be further shortened, and the responsiveness during step-down can be improved.

[0065] 7 is a flowchart showing a specific example of a flow rate control method in another embodiment. In this embodiment, too, the flow rate control operation by first-order delay control for performing step-down is performed by adjusting the current control value A while the normal flow rate control operation is being performed. n The control starts when a final target value B smaller than the reference value B is given, and is performed by the control loop shown in FIG.

[0066] As shown in FIG. 7, in the control loop employed in this embodiment, unlike the embodiment shown in FIG. 5, it is not particularly determined whether or not it is the first cycle of control, and as shown in steps S11 and S12, real-time update of the time constant τ is selected, and the time constant τ2 to be used is determined. n is calculated for each control cycle, for example, according to equation (4) above.

[0067] Other than that, similar to the flowchart shown in Figure 5, the control target value is updated by first-order delay control using the calculated τ2. Specifically, A n =A n-1 -(A n-1 The control target value is determined based on (τ2+1) / (τ2-B) (step S13), n is incremented (step S14), and then, as shown in step S15, the control target value A determined in step S13 is n The control valve 6 is subjected to feedback (FB) control in accordance with the above.

[0068] As described above, according to the flow control device and flow control method of the embodiment of the present invention, by performing first-order delay control of the control valve using the time constant τ2 corresponding to the final target value, it is possible to perform flow rate step-down with better responsiveness while preventing undershoot. [Industrial Applicability]

[0069] A flow rate control device and a flow rate control method according to an embodiment of the present invention are used to control the flow rate of gas supplied in semiconductor manufacturing facilities and the like. [Explanation of symbols]

[0070] 1 Flow path 2. Constriction section 3 Upstream pressure sensor 4 downstream pressure sensor 5 Temperature Sensor 6 Control valve 7 Control Mechanism 8 Flow Control Device 9. Shut-off valve 10. Process chamber 11 Vacuum pump 12 External control device

Claims

1. A flow rate control device comprising: a throttle portion; a control valve provided upstream of the throttle portion; an upstream pressure sensor that detects pressure between the throttle portion and the control valve; and a control mechanism connected to the control valve and the upstream pressure sensor, the flow rate control device being configured to control the control valve based on the output of the upstream pressure sensor, when an operation of closing the control valve is performed to reduce the flow rate of the gas flowing downstream of the throttle portion from an initial value A % to a final target value B %, the opening of the control valve is controlled based on an internal flow rate signal that updates the control target value at predetermined intervals, The internal flow signal is generated based on the following control equation, which includes a time constant τ: Control target value = previous target value + (final target value - previous target value) / (τ2 + 1) Here, the time constant τ2 is smaller than the time constant τ1 when decreasing from the initial value A% to the final target value 0%, and is set so that the curve defined by the control equation using the time constant τ2 does not fall below the curve defined by the control equation using the time constant τ1 at any time, and a flow control device, wherein the time constant τ1 is set to be larger than the time constant of a reference flow rate reduction characteristic, which is a flow rate characteristic when residual gas between the control valve and the throttling portion flows through the throttling portion after the control valve is closed.

2. 2. The flow control device according to claim 1, wherein the time constant τ2 is a fixed value over a period in which the gas flow rate is reduced from an initial value A % to a final target value B %, and the time constant τ2 is determined by the following formula: τ2=τ1×(AB) / AB / A

3. 2. The flow control device according to claim 1, wherein the time constant τ2 is updated every time the control target value is updated at the predetermined interval during a period in which the gas flow rate is reduced from an initial value A % to a final target value B %.

4. A flow rate control method executed in a flow rate control device including a throttle section, a control valve provided upstream of the throttle section, an upstream pressure sensor that detects pressure between the throttle section and the control valve, and a control mechanism connected to the control valve and the upstream pressure sensor, receiving a signal to reduce the flow rate of gas flowing downstream of the restriction from an initial value A% to a final target value B%; generating an internal flow signal for controlling the opening of said control valve upon receiving a signal to reduce flow; feedback controlling the opening of the control valve based on the output of the upstream pressure sensor so as to follow the internal flow rate signal; Including, The internal flow rate signal is a signal that updates the control target value at predetermined intervals and is generated based on the following control formula including a time constant τ2: Control target value = previous target value + (final target value - previous target value) / (τ2 + 1) The time constant τ2 is smaller than the time constant τ1 when decreasing from the initial value A% to the final target value 0%, and is set so that the curve defined by the control equation using the time constant τ2 does not fall below the curve defined by the control equation using the time constant τ1 at any time, and a flow rate control method in which the time constant τ1 is set to be larger than a time constant of a reference flow rate decrease characteristic, which is a flow rate characteristic when residual gas between the control valve and the throttle portion flows through the throttle portion after the control valve is closed.

Images