Pirani vacuum gauge

Abstract

A thermally conductive vacuum gauge implements a power dissipation model to accurately measure gas pressure. The enclosure encloses a gas volume, and the sensor wire is positioned within the gas volume. The controller provides a power dissipation model of the thermally conductive vacuum gauge that includes power losses due to thermal conduction losses of the sensor wire end contacts, radiation losses from the sensor wire to the gas enclosure, and pressure-dependent thermal conduction losses from the sensor wire by surrounding gas. The controller then applies a power input to the sensor wire to heat the sensor wire and measures a total power dissipation WT, a sensor wire temperature Ts, and an enclosure temperature Te during application of the power input. The gas pressure within the enclosure is determined based on the measured WT, Ts, and Te and a power dissipation model.

Description

[0001] Related Applications

[0002] This application is a continuation of U.S. Application No. 18 / 045,685 filed on October 11, 2022. The entire teachings of the above application are incorporated herein by reference. Background Art

[0003] Since the rate of heat transfer through a gas is a function of the gas pressure, under certain conditions, with appropriate calibration, the measurement of the rate of heat transfer from a heated sensing component to the gas can be used to determine the gas pressure. This principle is used in the well-known Pirani vacuum gauge, in which the heat loss is measured using a Wheatstone bridge network or other circuit that is used to both heat the sensing component and measure its resistance. In a Pirani vacuum gauge, a temperature sensitive resistor is connected as one arm of the Wheatstone bridge. The temperature sensitive resistor is exposed to the vacuum environment in which the pressure is to be measured.

[0004] Conventional Pirani gauges are calibrated for several known pressures to determine the relationship between gas pressure and gas power loss or bridge voltage. Then, assuming tip and radiation losses remain constant, the unknown pressure of the gas can be determined directly from the power lost by the gas or related to the bridge voltage when the bridge is balanced. Summary of the invention

[0005] Example embodiments include a thermal conductivity vacuum gauge including an enclosure surrounding a gas volume with a sensor wire positioned within the gas volume and a controller. The controller may be configured to provide a power dissipation model of the thermal conductivity vacuum gauge that includes power losses due to conductive heat loss at the sensor wire end contacts, radiation losses from the sensor wire to the gas enclosure, and pressure-dependent conductive heat loss from the sensor wire through the surrounding gas. The controller may then apply a power input to the sensor wire to heat the sensor wire and measure the total power dissipation W during the application of the power input. T , sensor wire temperature T s , and the shell temperature T e The controller can further be based on the measured W T , T s , and T e and power dissipation model to determine the gas pressure inside the enclosure.

[0006] The thermal conductivity vacuum gauge may be a Pirani vacuum gauge. The controller may be further configured to: 1) at a subsequent time, measure the total power dissipation W during the application of the power input T , sensor wire temperature T s , and the shell temperature T eand 2) determining a change over time of at least one of the end loss coefficient G and the radiation loss coefficient E based on the subsequent value.

[0007] The controller can be further configured to: 1) measure the total power dissipation W when the power input is different T and the sensor wire temperature T s 2) Determine the total power dissipation W T and the sensor wire temperature T s 3) Determine the values ​​of the end loss coefficient G and the radiation loss coefficient E based on the mathematical fitting. When the power input is different, the enclosure temperature T e The controller may be further configured to output a notification to remove and replace the sensor wire based on a comparison of at least one of the end loss coefficient G and the radiation loss coefficient E with a reference value. The controller may be further configured to 1) determine a gas adaptation coefficient A based on a mathematical fit, the gas adaptation coefficient A depending on the type of gas within the housing; and 2) determine a measure of gas pressure within the housing based on the gas adaptation coefficient A.

[0008] The controller can be further configured to 1) based on the sensor wire temperature T s The change in the sensor wire C over a given period of time determines S and 2) based on the sensor wire C S The controller can determine the sensor wire C based on the cooling rate of the sensor wire temperature Ts within a given period of time. S The controller may determine a measure of gas pressure during a period of increased gas pressure within the enclosure.

[0009] Additional embodiments include a method of operating a thermal conductivity vacuum gauge including a sensor wire in a gas volume within an enclosure. A power dissipation model for the thermal conductivity vacuum gauge may be provided, the power dissipation model including power losses due to heat conduction losses at the sensor wire end contacts, radiation losses from the sensor wire to the gas enclosure, and pressure-dependent heat conduction losses from the sensor wire through the surrounding gas. A power input may be applied to the sensor wire, and the total power dissipation W during the application of the power input may be measured. T , sensor wire temperature T s , and the shell temperature T e . Then, based on the measured W T , T s , and T e and power dissipation model to determine the gas pressure inside the enclosure.

[0010] The zero offset of a thermal conductivity vacuum gauge can be modeled by: 1) evacuating the enclosure to essentially zero pressure; 2) applying a power input to the sensor wires; 3) measuring the total power dissipation W during the applied power input. T , sensor wire temperature T s , and the shell temperature T e ; and 3) according to W T , T s and T e A tip loss coefficient G and a radiation loss coefficient E in a power dissipation model are determined. The tip loss coefficient G may correspond to heat loss at a terminal post coupled to the sensor wire during application of power input to the sensor wire. The radiation loss coefficient E may correspond to radiation loss of the sensor wire during application of power input to the sensor wire. A temperature compensation value may be determined to compensate for changes in at least one of a sensor wire temperature and a housing temperature, the temperature compensation value being a function of at least one of the tip loss coefficient G and the radiation loss coefficient E; and a gas pressure measurement may be determined based on the temperature compensation value.

[0011] Another embodiment includes a method of operating a thermal conductivity vacuum gauge. A power input may be applied to the sensor wire when the sensor wire and the housing housing the thermal conductivity vacuum gauge exhibit substantially zero pressure. The total power dissipation W during the application of the power input may be measured. T , sensor wire temperature T s , and the shell temperature T e . Can be based on the sensor wire temperature T s The change in the sensor wire C over a given period of time determines S Based on the heat capacity of the sensor wire C S and a measure of the power input applied to the sensor wires determine a measure of gas pressure.

[0012] The heat capacity C of the sensor wire can be determined based on the cooling rate of the sensor wire temperature Ts within a given period of time. S A measure of the gas pressure during the period of increased gas pressure within the enclosure may be determined.

[0013] Another embodiment includes a thermal conductivity vacuum gauge including a housing and a controller, the housing enclosing a gas volume within which a sensor wire is positioned. The controller can be configured to: 1) apply a power input to the sensor wire when the housing containing the sensor wire and the housing exhibits substantially zero pressure; 2) measure the total power dissipation W of the sensor wire when the sensor wire is heated by the application of the power input; T , sensor wire temperature T s , and the shell temperature T e ; 3) Based on the sensor wire temperature Ts The change in the sensor wire C over a given period of time determines S and 4) based on the heat capacity of the sensor wire C S and a measure of the power input applied to the sensor wires determine a measure of the gas pressure within the enclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The foregoing will be apparent from the following more particular description of example embodiments, as illustrated in the accompanying drawings, wherein like reference numerals refer to the same parts throughout the different views. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating the embodiments.

[0015] FIG. 1A is a circuit diagram of a prior art Pirani vacuum gauge.

[0016] FIG. 1B is a graph illustrating the response of the Pirani vacuum gauge of FIG. 1A .

[0017] 2A and 2B illustrate a prior art Pirani vacuum gauge including a compensation wire.

[0018] FIG. 3 illustrates the prior art Pirani vacuum gauge of FIGS. 2A and 2B implemented within a chamber.

[0019] FIG. 4 illustrates a sensor of a prior art thermal conductivity vacuum gauge used in an example embodiment.

[0020] Figure 5A The sensor of FIG. 4 is illustrated in further detail.

[0021] Figure 5B is a graph illustrating the response of the sensor of FIG. 4 .

[0022] Figure 6 is a graph illustrating the effect of "zero pressure" losses on overall heat dissipation in a sensor in one embodiment.

[0023] Figure 7 is a graph depicting power dissipated at zero pressure versus wire temperature in one embodiment.

[0024] Fig. 8A and Figure 8B is a graph depicting power consumption at a sensor of a vacuum gauge in one embodiment.

[0025] Fig. 9 is a diagram depicting the interaction between heated sensor wire and gas molecules.

[0026] Fig. 10A and Fig. 10B is a graph of power as a function of pressure in one embodiment.

[0027] Fig.11 is a block diagram of a vacuum gauge circuit in one embodiment.

[0028] Fig.12 is a graph of normalized pressure versus time in one embodiment.

[0029] Fig.13 and Fig.14 is a graph of sensor temperature versus time in one embodiment.

[0030] FIG. 15A to FIG. 15D is a graph of sensor temperature versus time in another embodiment.

[0031] Fig.16 is a graph of (1 / T) versus P in one embodiment.

[0032] Fig.17 is a graph of heat capacity as a function of wire temperature.

[0033] Fig.18 is a graph of normalized pressure versus time in one embodiment. DETAILED DESCRIPTION

[0034] A description of example embodiments follows.The teachings of all patents, published applications, and references cited herein are incorporated by reference in their entirety.

[0035] Example embodiments include thermal conductivity vacuum gauges (such as Pirani vacuum gauges) and methods of operating the same. A conventional Pirani vacuum gauge is described below with reference to FIGS. 1A to 3 , and an additional vacuum gauge is described below with reference to FIG. 4 . Such a vacuum gauge may be implemented below with reference to FIG. 5A to FIG. 18 In the example implementation described.

[0036] Pirani gauges with constant sensor wire temperature have been used to perform pressure measurements between 1E-4 Torr and 1000 Torr. A typical Pirani gauge that provides a constant sensor wire temperature during operation relies on a Wheatstone bridge connected to the sensor wire. The electrical power required to keep the wire at a constant temperature is used to provide a pressure measurement. Maintaining a constant temperature at the sensor wire is desirable because it allows for a faster response to pressure steps without having to wait for a temperature change to occur. In addition, having a constant wire temperature provides a pressure-independent signal baseline offset that can be subtracted from the actual signal to provide the pure pressure-related portion of the signal itself.

[0037] In a typical constant wire temperature Pirani gauge, the temperature of the wire is maintained at a constant temperature by passing pressure-dependent electrical heating power through the wire. Since the electrical power required to maintain the wire at a constant temperature depends on the pressure, a simple power measurement is used to provide a pressure measurement. This design relies on the Wheatstone bridge to maintain its temperature-dependent resistance during operation to regulate the wire temperature.

[0038] FIG1A is a circuit diagram of a prior art Pirani vacuum gauge 100. The pressure sensor includes a temperature sensitive resistor R connected as one arm of a Wheatstone bridge 110. S . R 3 Typically a temperature sensitive resistor, designed to 3 The temperature rise caused by R 2 and R 1 Typically a fixed resistor. Sensor wire R S and typically R 3 The sensor is exposed to the environment where the pressure is to be measured. The environment is contained within the enclosure and the sensor lead wire R S Alternatively, R 3 It may also be included in the enclosure via the addition of one or more feed-throughs.

[0039] The resistance value of the resistor R 1 , R 2 and R 3 is chosen so that when the pressure-dependent voltage V 电桥 When applied to the top of the bridge, V 左 =V 右 , sensor wire R S The resistance is fixed and equal to (R 1 *R 3 ) / R 2 Voltage V 电桥 The operational amplifier automatically controls V 左 With V 右 The voltage difference between them is maintained at zero volts. 左 To V 右 The bridge is considered balanced when the potential drop across it is zero. When the bridge is balanced, the following conditions exist:

[0040] a) is = i3, (1)

[0041] b) i1 = i2, (2)

[0042] c) isRS = i1R1, (3)

[0043] d) i2R2 = i3R3, (4)

[0044] e) Dividing Equation 3 by Equation 4 and using Equations 1 and 2 gives

[0045] f) RS = βR3 (5), where β = R1 R2 (6)

[0046] g) Therefore, when the bridge is balanced, RS is a constant fraction β of R3.

[0047] h) To achieve steady-state conditions for RS at any given pressure, the following equation 7 must be satisfied:

[0048] i) Electrical power input to RS = power radiated by RS + power lost at the end of RS + power lost to gas by RS (7)

[0049] Because the sensor resistor R S The amount of electrical power required to maintain a constant temperature and constant resistance increases with pressure, so the voltage V 电桥 also depends on pressure. This relationship is illustrated in Figure 1B, which is at R S The pressure range within the occupied chamber is the voltage V 电桥 As shown in the figure, the voltage V 电桥 The pressure range shows an S-shaped curve. Conventional Pirani vacuum gauges are calibrated based on several known pressures to determine the unknown pressure P x The relationship between the power loss to the gas and the power loss of the gas, or more conveniently, the relationship with the bridge voltage. Then, assuming that the end losses and radiation losses remain constant, the unknown pressure of the gas, P x It can be determined directly from the power lost to the gas or can be related to the bridge voltage when the bridge is balanced.

[0050] The Pirani vacuum gauge 100 provides a simple configuration for measuring pressure and allows adjustment of the sensor lead resistance in a simple manner. A simple op amp circuit can be used to zero the bridge (V 左 =V 右 ), allowing for low-cost construction of the circuit. However, in order to provide compensation for different ambient temperatures outside the chamber, resistors with highly specific values ​​must be added to the measurement head during the calibration process to provide the desired signal response (i.e., V 电桥 relative to pressure) and appropriate temperature dependence.

[0051] 2A and 2B illustrate a prior art Pirani vacuum gauge 200 including a compensation wire Rc. The vacuum gauge 200 is comparable to the Pirani vacuum gauge 100 described above, but the addition of the compensation wire Rc allows the vacuum gauge 200 to compensate the pressure reading for ambient temperature fluctuations. Such ambient temperature fluctuations change the sensor wire Rc. SThe temperature difference between the temperature of the enclosure wall (not shown) containing the chamber whose pressure is to be measured. As shown in FIG2B, the compensation wire resistor R C Wrapped around a smaller envelope within the chamber and allowed to reach a temperature T in thermal equilibrium with room temperature 2 Then tune the resistance in the bridge (R 3 and R 4 ) and the resistance in the compensation wire Rc, so that T 2 changes, and when the Wheatstone bridge is balanced, the temperature difference T 1 -T 2 (where T 1 is the sensor R S The wire temperature of the sensor wire R S The power difference dissipated into the gas depends on this temperature difference, so measurement of this power dissipation makes the pressure measurement independent of the ambient temperature.

[0052] In practice, the compensation wire R C Therefore, each embodiment of the vacuum gauge 200 must be individually tuned by adjusting the resistance value during testing and calibration to provide a temperature differential (T) that remains constant as the ambient temperature changes. 1 -T 2 ). Further, the compensation wire R C Winding can be expensive and difficult to accomplish. To provide fast response, the compensating wire R C It can also be wound inside a vacuum gauge in a thin-walled enclosure and exposed to the gas environment.

[0053] FIG. 3 shows the above-described prior art Pirani vacuum gauge 200 in another view implemented within a chamber 290 (not shown to scale). A portion of the vacuum gauge 200, including the sensor wire R S and compensation wire R C , extends into the cavity 290 through the feedthrough flange 220, while the rest of the Wheatstone bridge remains outside the cavity 290. C The pressure sensor volume installed on the thin-walled tank 240 is conducive to compensating the wire R when the room temperature changes. C The vacuum gauge 200 requires at least four feedthroughs 210 that pass through the feedthrough flange: two feedthroughs connect the sensor wires R S , and the other two feedthroughs are connected to the compensation wire R C .

[0054] The Pirani vacuum gauge 200 exhibits several disadvantages. In particular, calibration of the vacuum gauge 200 can be difficult and laborious. The vacuum gauge 200 must be calibrated to obtain proper temperature compensation, including selecting appropriate resistor values ​​and ensuring that the value T1 -T 2 remains constant regardless of room temperature. The Wheatstone bridge requires fine tuning for temperature compensation. If the calibration procedure is performed correctly, a maintained value T can be achieved 1 -T 2 , but does not allow the use of nominal resistor values. Instead, each vacuum gauge must be manually tuned and configured with specific resistors as high-precision components.

[0055] Due to the strict implementation of Wheatstone bridge based temperature control, the vacuum gauge 200 does not allow the sensor wire operating temperature (or resistance) to vary during operation, but rather provides a single operating temperature.

[0056] Even if the pressure is maintained with the sensor wire R S There is a linear relationship between the power required to maintain a constant temperature and the vacuum gauge 200 also indicates that the vacuum gauge 200 is based on the bridge voltage V as shown in FIG. 1A. 电桥 The measured pressure is not linearly related to pressure. Large baseline offsets (e.g. due to radiation losses and tip losses) are associated with V 电桥 The non-linear response to pressure produces an S-shaped curve, which makes calibration difficult and less accurate when interpolating measurements.

[0057] FIG4 shows a thermal conductivity vacuum gauge 400 in another embodiment. Note the sensor portion of the vacuum gauge. The vacuum gauge 400 includes a sensor wire R S 405 (also referred to as a filament) is secured within chamber 490 by wire mount 406. Wire 405 is connected to vacuum gauge circuitry 450 via terminals 412, which extend into chamber 490 via a single feedthrough 410 (e.g., electrically connected via an air-to-vacuum feedthrough connection). The opposing node of wire 405 can be grounded, such as enclosure 480 containing chamber 490. Temperature sensor 470 (e.g., a thermistor) can be positioned at or near enclosure 480 to measure the temperature of enclosure 480 and / or the ambient temperature outside chamber 490.

[0058] Compared to the vacuum gauge 200 described above with reference to Figures 2A to 3, the vacuum gauge 400 provides a sensor with a simpler configuration. The vacuum gauge 200 only requires a single feedthrough 410 that enters the chamber 490. Further, the compensation wire can be omitted from the vacuum gauge 400 because the temperature compensation can be provided using the temperature sensor 470 in conjunction with the vacuum gauge circuit 450. Therefore, the vacuum gauge 400 enables the sensor to have a simpler, more compact structure, and less labor is required to assemble. The vacuum gauge 400 and other embodiments of the thermal conductivity vacuum gauge that can be used in the example embodiments are described in U.S. Patent No. 10,845,263, and the entire teachings of the patent are incorporated herein by reference.

[0059] Many conventional thermal conductivity vacuum gauges utilize a Wheatstone bridge circuit as described above to (1) control the sensor wire temperature and (2) access the pressure-dependent bridge voltage (Vb) being measured through a factory-based calibration procedure. Several vacuum gauges rely on constant sensor wire temperature and fast analog temperature control for real-time pressure tracking. Although measurement of the bridge voltage is a common calibration approach in such vacuum gauges, this approach does not provide the opportunity to monitor or track the thermal characteristics of the sensor during manufacturing and operation. Pirani sensor pressure readings are typically sensitive to ambient temperature changes and are expected to operate over a wide range of ambient temperatures.

[0060] Conventional vacuum gauges handle ambient temperature compensation in one of two ways: (1) temperature-sensitive compensation resistors added to the bridge (i.e., Vb is temperature independent during assembly), or (2) temperature correction coefficients determined during a test / calibration process in a constant temperature test vacuum system (e.g., a temperature and / or pressure controlled environment). A typical pressure calibration procedure generates or simply verifies a lookup table of bridge voltage (Vb) versus pressure (P), which is then stored in the vacuum gauge / transducer electronic control unit. Sensor lead temperature is typically not included in the pressure calculation because the lead temperature is controlled but not actively or accurately measured during operation. Case temperature is sometimes used for temperature correction of pressure readings. The calibration coefficients used for temperature correction are typically measured over a narrow pressure range in a constant temperature test vacuum system during a lengthy procedure, and do not provide adequate temperature compensation over the entire pressure range of commercial sensors because they do not properly account for the different heat loss mechanisms that exist throughout the operating pressure range. In addition to changes in case temperature due to changes in ambient temperature, self-heating due to higher power dissipation at higher pressures can also cause the case to operate above ambient temperature as discussed later. Some commercial products focus on temperature correction capability to minimize the effects of self-heating on the pressure display.

[0061] The example embodiments described below may implement processes for manufacturing and / or calibrating thermal conductivity vacuum gauges that improve the calibration and operation of the gauges. Such embodiments may implement sensor hardware such as described above.

[0062] Figure 5AA thermal conductivity vacuum gauge 500 in an example embodiment is illustrated. The vacuum gauge 500 may incorporate features of a thermal vacuum gauge disclosed in the above-cited U.S. Patent No. 10,845,263. The sensor is configured to add a shield 515 as needed. The sensor wire 505 may be connected between a terminal 512 (implemented as a feedthrough pin) and the shield 515. The shield 515 provides a conductive path to ground and surrounds at least a portion of the sensor wire 505, thereby protecting the sensor wire 505 from physical damage from contaminants in the process environment and providing thermal boundary conditions for the sensor wire 505. When used in conjunction with a hot cathode vacuum gauge, the shield 515 may also be used to shield the sensor wire from radiation from a hot filament. In such a configuration without the shield 515, the sensor wire may experience a large baseline radiation offset change. An insulator 511 may surround the terminal 512 at the feedthrough 510 to ensure a seal within the chamber 590 contained by the enclosure 580. The terminal 512 is further connected to the controller 550.

[0063] The sensor wire 505 can be a thin wire of small diameter (e.g., 0.001 inch or 0.002 inch) and twisted into a coil (e.g., a coil with a diameter of 0.010 inch). The operating temperature T1 of the sensor wire 505 can be selected to have a target of 20°C or more above room temperature to provide sufficient sensitivity to pressure changes. The temperature of the sensor wire 505 can be maintained at a constant value during operation, which can improve the speed of response to pressure changes. The controlled power input (specified by W) can be controlled to a certain value. T This constant temperature T1 is achieved by applying a pressure (different from the pressure P) to the terminal 512 to bring the sensor wire 505 close to the target resistance value. For the sensor wire 505, the relationship between the resistance and temperature of the sensor wire 505 can be determined based on previous measurements of the same wire type. This relationship can be used to calibrate the vacuum gauge 500. Figure 5B As shown, the required power input W T also varies depending on the pressure of chamber 590. This function exhibits a linear region where pressure can be measured most accurately. For vacuum gauges using a bridge circuit, the bridge can fix the operating temperature of the wire, and although the temperature may not be measured, it can be assumed to be constant. However, this assumption may not always hold, especially during pressure transients.

[0064] To improve the operation and calibration of thermal conductivity vacuum gauges, example embodiments may utilize real-time measurements of multiple measurands, including total power dissipation (W T ), sensor wire temperature (T s ) and the shell temperature (T e). The heat loss mechanisms of Pirani sensors are well understood and applicable thermal models are widely available from the technical literature. However, conventional Pirani vacuum gauges have not implemented any such thermal or physical models to improve the manufacturing and measurement capabilities of Pirani vacuum gauges. The manufacturing, calibration, and pressure measurement of thermal conductivity vacuum gauges can benefit from thermal models in several ways as described below.

[0065] Total power dissipation W T represents the amount of heat transferred from the sensor wire to the surrounding enclosure through several mechanisms including: end losses of gas molecules, radiation losses, and thermal conductivity - it is dependent on the gas pressure and is parameterized by both the sensor wire and enclosure temperatures. According to a well-established thermal model, W T Increases in direct proportion to the gas pressure and the temperature difference between the sensor wires and the enclosure wall. The detailed mathematical equations relating power dissipation to gas pressure and temperature are listed below. Gas Power Dissipation (W g ) can be isolated from the total power consumed by the sensor wire during operation and is based on the known thermophysical properties of heated sensor wires immersed in a gas. The gas power loss is proportional to the thermal conductivity of the gas and proportional to the gas pressure. When the gas power loss is known, the accuracy of the pressure calculation and the efficacy of temperature compensation are improved. Pirani sensors are also described as thermal conductivity sensors, which provide an indirect measurement of pressure, i.e., dependent on the chemical composition of the gas. Mathematically, W g With W T The ability to isolate simplifies pressure calculations, improves temperature correction algorithms, provides the ability to calculate pressure for other gases (i.e., gases different from the calibration gas), and provides a path for predictive maintenance.

[0066] Sensor wire temperature T s It can be measured based on the resistance of the sensor wire, which indicates the operating temperature of the sensor wire and affects the total power dissipation W T and gas power dissipation W g Without sensor wire temperature measurement, a Pirani gauge may not be able to distinguish between power changes due to pressure or wire temperature fluctuations. Wire temperature control (analog or digital) is never perfect, and wire temperature may change from moment to moment due to pressure transients. Real-time sensor wire temperature T s The measurement enables example embodiments to isolate gas power dissipation variations. Sensor wire temperature measurement is also necessary to provide accurate (as described below) temperature compensated "zero pressure" power measurements and to derive temperature coefficients without the need for a constant temperature test vacuum system.

[0067] The power dissipation mechanisms that drive heat loss away from the heated wire are different in different pressure ranges. -3At a pressure of 10°, heat dissipation is dominated by radiation and end losses, since gas conductivity losses are negligible. -3 In the range of 10 torr to 10 torr, heat transfer is strictly related to the variation of gas thermal conductivity with pressure. Finally, above 10 torr, heat loss is dominated by gas conduction, but is also affected by rising gas temperature and increased convection. Each of these heat loss mechanisms is described in further detail below.

[0068] "Zero pressure" loss (high vacuum range, P < 1E-6 Torr)

[0069] As used herein, "zero pressure" refers to the absence of gas density detectable by the thermal conductivity vacuum gauge. In the absence of detectable gas density, heat dissipation in the sensor wire is dominated by two independent mechanisms: tip losses and radiation losses. At high vacuum levels, the total power loss W T It can be expressed as:

[0070] W T = W 端部 + W 辐射 (8)

[0071] in:

[0072] W 端部 = the power dissipated to the end legs (i.e., end losses), and

[0073] W 辐射 = power dissipated by radiation (i.e., radiation loss).

[0074] End posts may refer to posts that are directly connected to each end of a sensor wire (e.g., wire mount 406 as shown in FIG. 4). 端部 and W 辐射 The relative contribution of is affected by the surface emissivity properties, the bulk thermal conductivity of the sensor wire, the dimensions of the wire (eg, length and diameter), the operating temperature of the wire, and the operating temperature of the enclosure.

[0075] Figure 6 is a graph illustrating the effect of zero pressure loss on overall heat dissipation in an example sensor. The graph highlights the observation that zero pressure loss is typically determined by the end losses of the sensor wire at the normal reference operating temperature (i.e., T ref ≈100°C, typical) dominates, the contribution of radiation is low, provided the sensor wires are kept clean and polished. Because the contribution of radiation losses to zero pressure loss decreases with T s 4 As the sensor wire temperature increases, the radiation losses are expected to increase relative to the tip losses (as T s Increase).

[0076] End loss: W 端部

[0077] During operation, the terminal block dissipates heat to the sensor housing and maintains a temperature close to the ambient temperature T 环境 As a result, a constant thermal power transfer flux W occurs from the wire guide pin. 端部 The end posts are thermal anchors with large heat capacity (or thermal mass) and thermal conductivity to the enclosure. The heat dissipated to the end posts is supplemented by an electrical heater circuit to maintain a constant wire temperature (e.g., a constant resistance Rs value).

[0078] In the thermal model, the tip losses can be mathematically expressed as:

[0079] W 端部 = G * (T s – T e ) (9)

[0080] in:

[0081] G = End Loss Coefficient (ELC)

[0082] T s = Wire temperature (derived from R s )

[0083] T e = Shell temperature

[0084] Ambient temperature T 环境 and the shell temperature T e The difference between T and T is introduced into equation (9) because when self-heating begins at the highest pressure, the temperature of the enclosure (to which the sensor wires are dissipated) may rise above the ambient temperature. e It will continue to be used throughout the thermal analysis to account for possible differences between the enclosure temperature and the ambient temperature due to self-heating.

[0085] The above function form shows that W T With T s The linear relationship between indicates that the total heat dissipation is dominated by the tip losses. ELC can be a function of: 1) sensor lead material (substrate lead and coating), 2) sensor lead size, 3) post size, 4) post material, 5) post to header thermal connection, and 6) notch characteristics. If the mechanical characteristics of the sensor do not change over time (i.e., material does not consume or accumulate), then ELC(G) can be expected to remain fairly constant. It can also be expected that ELC is repeatable from unit to unit.

[0086] Radiation loss: W 辐射

[0087] The heated sensor emits electromagnetic radiation to its surrounding structure (blackbody emission). The rate of energy loss due to this process is proportional to 1) the temperature of the wire (non-uniform distribution), 2) the emissivity of the wire surface material, and 3) the temperature of the surrounding structure (roughly T e ). In the thermal model, the radiation loss can be expressed mathematically as:

[0088] W 辐射 =E*(T s 4 –T e 4 )(10)

[0089] in:

[0090] E = Radiation loss coefficient (RLC)

[0091] T 参考 = Reference temperature of the conductor (derived from Rs)

[0092] T e = Shell temperature

[0093] The above equation shows the nonlinear relationship between radiated power and temperature (eg, absolute temperature). RLC is proportional to the area and emissivity coefficient of the sensor wire surface.

[0094] Figure 7 Depicted in T e = The power dissipated at zero pressure of the sensor at 20°C relative to T s Because T s When reaching high temperature values ​​in this experiment, the influence of radiation losses can be clearly noticed, that is, with T s As the value of increases, the slope increases upward. Fit the power of the above curve to the functional form:

[0095] W T (zero pressure) = G*(T s –T e )+E*(T s 4 –T e 4 )(11)

[0096] The coefficients derived from the nonlinear fit indicated by the traces are shown in Table 1:

[0097] dissipation coefficient value End loss (ELC) G 4.36e-06W / K Radiation (RLC) E <![CDATA[3.85e-15W / K 4 ]]>

[0098] Table 1

[0099] The above measurements and calculations offer several advantages. In particular, W T , T s and Te The measurement of enables the calculation of the G and E thermal coefficients for each vacuum gauge during (1) the calibration process or (2) sensor operation (i.e., each time the sensor is operated at zero pressure and stabilized). The validity of equation (11) can be verified experimentally and it is shown that tracking power as T s or T e The same value of the G and E coefficients can be obtained by changing , as depicted in equation (11).

[0100] Based on equation (11), during the manufacturing / calibration process, the s or T e to determine the G and E coefficients. Pirani circuit implementations, such as the example implementation vacuum gauge described below, allow the sensor wire temperature to vary and can generate new G and E values ​​in seconds (both at the factory and in the field). In contrast, deriving G and E from Te changes requires a constant temperature test vacuum system and is an expensive and time consuming process. In the factory or in the field, the calculation of G and E is particularly simple in the example implementation because it allows the sensor wire temperature to vary over a wide range: the total power can be measured at multiple wire temperatures, and G and E can be calculated to fit the functional form of equation (11) while remaining in a high vacuum. In the absence of variable wire temperature capability, measurement of G and E can be achieved by varying Te in a constant temperature test vacuum system, which is a much slower process.

[0101] Furthermore, under zero pressure conditions, W T , T s , T e The measurement of can determine the thermal coefficients: G and E, which provides several advantages. During the manufacture of vacuum gauges, determining G and E at the factory can enable real-time incoming material identification and improve manufacturing yields, preventing unqualified materials from entering the production floor. In an example embodiment, the measurement of end loss and emissivity coefficients can provide a means to establish an incoming material inspection program and preventive fault reporting. Such an embodiment can implement full digital power control and power and temperature measurement, or a hybrid configuration including an analog temperature controller can be implemented. Such an embodiment can also measure power and temperature in real time, including a configuration with a bridge as described above, which can be accomplished by adding additional test points to the bridge circuit to measure the current entering the sensor wires and the voltage on the sensor wires. As part of the statistical process control (SPC) analysis of incoming material characteristics, the G and E coefficients can be tracked during the production process.

[0102] The measurement also provides operational advantages. Determine G and E in situ (using the sensor’s variable T sThe vacuum gauge may be configured to detect sudden increases in E associated with contamination (which affects minimum detectable pressure performance) or sudden changes in G that may be due to corrosion or contaminant buildup on the sensor wires and may indicate that the useful life is about to end (and preventive maintenance is needed). The G and E coefficients may be tracked during operation to provide predictive maintenance prompts, as described in further detail below. In addition, the use of G and E and the ability to upgrade the thermal coefficient of the vacuum gauge in the field provides improved zero pressure measurement and minimum detectable pressure (MDP) specifications that are less dependent on ambient temperature. Tracking G and E is an important feature for improving zero pressure measurement specifications in example embodiments.

[0103] Gas loss

[0104] As the gas pressure around the sensor wires begins to increase, the gas molecules begin to contribute to the total power loss. The total power dissipated is now expressed as:

[0105] W T = W 端部 + W 辐射 + W 气体 (12)

[0106] in:

[0107] W 端部 = Power dissipated to the terminal (end loss)

[0108] W 辐射 = Power dissipated by radiation (radiation loss)

[0109] W 气体 = Power consumed by gas (gas thermal conductivity loss)

[0110] The energy transfer from the heated wire to the gas molecules is pressure dependent; a Thermal Conductivity Gauge (TCG) uses this effect to derive the gas pressure. Since the heat transfer to the gas molecules is species dependent (heat capacity, mass, and accommodation coefficient), all Thermal Conductivity Gauges are essentially indirect pressure measuring sensors. According to Equation 12, and in order to determine the amount of power transferred to the gas molecules, the zero pressure power loss must be known and subtracted from the total power. Accurate pressure measurement requires "knowing" the zero pressure power loss at all times, which in turn requires knowing the G and E of the sensor to adjust for changes in wire or enclosure temperature. Knowing G and E allows for the measurement of pressure at T s or T e Subtract exactly zero when changing.

[0111] Fig. 8A is plotted for a given vacuum gauge and at several different housing temperatures between 0°C and 60°C and at T ref= Power dissipation at the sensor at 100°C. As shown, the power dissipation over the entire pressure range decreases with increasing package temperature. It is also clear that the variation of power with temperature depends on the pressure range.

[0112] Figure 8B It depicts the e =20℃, power dissipated by gas W 气体 As a function of pressure. By subtracting the zero pressure W from the total pressure data T To calculate the power consumed by the gas. Gas power, W 气体 There is a strict linear relationship between P and P for P < 1 Torr, and the slope of this linear response is proportional to the adaptation coefficient of the gas.

[0113] Low pressure operation – P ≤ 1 Torr

[0114] Fig. 9 is a diagram depicting the interaction between a heated sensor wire and gas molecules. At low pressure, the gas molecules collide with the surface of the hot wire. All molecules arrive at an average temperature of T g (and its corresponding Maxwell energy distribution). For low pressures (i.e., low pressures where the mean free path of molecular collisions is greater than the characteristic size of the sensor), the temperature of the gas molecules reaching the wire is in equilibrium with the enclosure temperature, T g =T e If the self-heating is not in place, T e =T 环境 Molecules that reach the wire typically spend a short residence time on the surface (related to the accommodation coefficient) and leave the filament at an elevated temperature close to the wire temperature.

[0115] The adaptation coefficient is typically defined as the probability of an energy transfer process. At low pressure, the heated molecules ejected from the wire collide with the enclosure wall many times, and the temperature of the molecules equilibrates to T before another collision with the sensor wire occurs. g =T e , energy is efficiently transferred from the sensor wire to the wall. The net power transferred is the gas power dissipation.

[0116] In the thermal model, the power dissipated by the gas at these low pressure conditions is expressed as:

[0117] W g = A * [(T s – T g ) / T g 1 / 2 ] * P (13)

[0118] in:

[0119] W g= power dissipated by gas molecules

[0120] A = Gas coefficient – ​​depends on the type. Includes adaptation factor.

[0121] T s =Sensor wire temperature – reference temperature value.

[0122] T g＝ The temperature of the gas molecules reaching the sensor wires.

[0123] At low pressure (i.e., in the absence of self-heating), T g =T amb =T e

[0124] P = Gas pressure

[0125] Equation (13) shows that at low gas density values ​​W 气体 The linear relationship between P.

[0126] Fig. 10A is a curve graph, which proves that in the case corresponding to N 2 The "gas loss" trace of the gas has a linear relationship as shown in equation (13) when the pressure is < 1 Torr. Fig. 10B A similar relationship is shown for data extracted using a power measurement circuit. The coefficient a can be calculated from the linear region of the graph using equation (13) after measuring: s , T e (=T 气体 ), and W T Table 2 shows the operation using multiple T e and T s Set the value of A to be calculated according to equation (13):

[0127]

[0128] Table 2

[0129] Using equation (13), at a typical T s and T e Under this combination, a consistent value of A can be calculated and used to calculate W in the linear region. g .

[0130] Therefore, the measurement under zero pressure conditions (W T , T s , T e ) can determine the thermal coefficient A (also called the gas adaptation coefficient). This determination can provide several advantages. For example, measuring A during the manufacturing process can provide the sensor wire for N 2A measure of the adaptability factor of a gas. Monitoring the factor a can become part of the incoming material identification process and prevent unsuitable (e.g. dirty or poor surface quality) materials, especially for use as sensor wires, from entering the production floor. In vacuum gauge calibration, a consistent factor a is valid for a wide range of pressures and temperatures, allowing the use of (W T , T e , T s ) and calculate low pressures over a wide pressure range (e.g., <1 Torr) from a single A value stored in the electronic control module. The coefficient a can be calculated from a single test pressure measurement, eliminating the need to use multiple pressure set points (e.g., reducing cycle time). When the coefficient a is available, low pressure calibration and measurements can be simplified.

[0131] Additional improvements are achieved in the measurement process. For example, using equation (13), a single stored coefficient a, the total power and the measured temperature T e and T s to calculate the temperature compensated pressure. Equations (11) and (13) provide specific temperature coefficients for temperature correction of zero pressure power and gas power. Relying on thermal models for all the different heat dissipation mechanisms and their thermal coefficients effectively extends the temperature and pressure range over which Pirani sensors can be temperature corrected. The temperature coefficients used are associated with verifiable physical phenomena and provide additional insight into the quality of the manufacturing materials and their health.

[0132] The exemplary embodiments described herein provide improvements in manufacturing, calibration, operation, and pressure measurement that can be obtained from thermal conductivity vacuum gauges. T , T s , T e Based on the thermal model of gas thermal conductivity as described above, example embodiments can characterize the thermal properties of the vacuum gauge and provide a faster and simpler calibration procedure, a more accurate pressure measurement algorithm, a wider range of temperature corrections for pressure readings, and improved tracking of sudden pressure transients (e.g., steps and pulses).

[0133] Shell temperature T e The measurement of can be performed using various thermometers that are thermally coupled to the enclosure wall. Thermistors, temperature diodes, platinum resistors are some examples of thermometers that are compatible with this application. The thermometer can be located on the vacuum side or the air side of the enclosure. The measurement of the electrical heating power delivered to the sensor can be achieved in a variety of ways through various circuit implementations.

[0134] Fig.11 1 is a block diagram of a vacuum gauge circuit 1100 that may be implemented in a thermal conductivity vacuum gauge in an example embodiment. Here, electrical heating power is delivered from a power supply 1110 to a sensor lead 1105, while a microprocessor 1120 controls the current (Is ) is delivered to the sensor lead 1105 to achieve the target sensor resistance R 0 Target resistance R 0 and the desired sensor wire temperature T s The digital measurement of the voltage Vs on the sensor lead by the microprocessor 1120 provides the sensor lead resistance R S and total power dissipation, since microprocessor 1120 knows I s .

[0135] During operation, the target sensor wire resistance R 0 To set the sensor wire 1105 temperature T s , the target sensor wire resistance corresponds to the desired temperature. In contrast to the Wheatstone bridge vacuum gauge implementation, T s The thermal coefficients G, E, and A can be changed by the microprocessor 1120 during calibration and in the field to derive or verify the thermal coefficients. The power supply 1110 is provided by I s Provide microprocessor 1120 to calculate W T The digital measurement of the voltage on the sensor wire 1105 provides the total power and wire resistance. The wire resistance can then be fed back to the temperature control loop and the resistance error signal used to close the loop. The temperature control loop can be fully digital and processed by the microprocessor 1120. The voltage measurement can be digital and a fast analog-to-digital converter with high resolution can be used at the microprocessor 1120.

[0136] Turn again Figure 5A , the controller 520 may implement some or all of the above-described measurement, calculation, and calibration features, and may be combined with the above-referenced Fig.11 The vacuum gauge circuit 1100 described above is shown in FIG. 1 . For example, the enclosure 580 may enclose a gas volume to be measured, and the sensor wire 515 may be positioned within the gas volume. The controller 520 may be configured to provide a power dissipation model for the thermal conductivity vacuum gauge 500 that includes heat conduction losses (e.g., W) due to the sensor wire end contacts. 端部 ), radiation losses from the sensor wires to the gas envelope (e.g., W 辐射 ) and the pressure-dependent heat loss from the sensor wire through the surrounding gas (W 气体 ) caused by the power loss. The controller 520 can then apply a power input to the sensor wire 515 to heat the sensor wire 515 and measure the total power dissipation W during the application of the power input. T , sensor wire temperature T s , and the shell temperature T e Based on the measured WT , T s and T e As well as the power dissipation model, controller 520 can determine the gas pressure within enclosure 580.

[0137] To provide predictive maintenance and field calibration, at a later time, the controller 520 may measure the total power dissipation W during the application of the power input. T , sensor wire temperature T S , and the shell temperature T e Based on the subsequent values, the controller 520 may determine a change in at least one of the end loss coefficient G and the radiation loss coefficient E over time.

[0138] The controller 520 can also measure the total power dissipation W when the power input changes T and the sensor wire temperature T s Then determine the total power dissipation W T and the sensor wire temperature T s Based on the mathematical fitting, the controller 520 can determine the values ​​of the end loss coefficient G and the radiation loss coefficient E. In this process, when the power input changes, the enclosure temperature T e Can be maintained at a constant value. The controller 520 can output a notification to remove and replace the sensor wire (or a larger component of the vacuum gauge that houses the wire) based on a comparison of at least one of the end loss coefficient G and the radiation loss coefficient E with a reference value. Based on mathematical fitting, the controller can also determine a gas adaptation coefficient A, which can depend on the type of gas in the housing 580, and then determine a measure of the gas pressure within the housing 580 based on A.

[0139] As described in further detail below, the controller 520 may also determine the heat capacity C of the sensor wire based on the change in the sensor wire temperature Ts over a given period of time. S , then based on the heat capacity C S A measure of the gas pressure within the enclosure is determined. The controller 520 may determine the heat capacity C based on the cooling rate of the sensor wire temperature Ts over a given period of time. S , and a measure of the gas pressure during the increase in gas pressure within the housing 580 can be determined.

[0140] When the above features are implemented, thermal conductivity vacuum gauges can exhibit several advantages. T , T s , T e Allows the controller to fully characterize and optimize the vacuum gauge based on a thermal model. Thermal coefficients, including but not limited to G, E, and A above, can be exported at the factory and updated in the field when the three measured quantities are available within the vacuum gauge.

[0141] Thermal coefficients generated at the factory during the calibration process can have several uses. For example, such thermal coefficients can be used to qualify incoming materials. Factory quality control will be able to set limits on acceptable values ​​for the thermal coefficients and reject unsuitable materials that do not meet such limits. Such quality control procedures ensure more consistent product performance for customers. Conventional sensor lead qualification procedures are often limited to lead resistance measurements, which is insufficient to ensure the most consistent unit-to-unit performance.

[0142] Thermal coefficients can also be used to provide improved minimum detectable pressure (MDP) performance. Pressure measurements at the lower end of the range depend on accurate recording of zero offset. Some conventional vacuum gauges already track zero offset, but no vacuum gauge is able to correct zero offset in real time as the sensor wire or ambient temperature changes. Zero offset drift limits the MDP specifications of all commercially available products. Therefore, users must update the zero offset of their instruments on a daily basis, and preferably each time zero pressure is reached (press a button on the unit, digital signal input, or command). Using the G and E coefficients will allow vacuum gauge users to reduce zero offset checks because the instrument will be able to update the zero offset based on sensor wire and enclosure temperature changes and based on an accepted thermal model of zero pressure power dissipation.

[0143] The thermal coefficient can be further used to simplify pressure calibration at low pressures. The ability to isolate gas power dissipation and use the A thermal coefficient to calculate pressure, with physics-based temperature compensation, allows for minimization of the number of test points required during low pressure calibration. The thermal coefficient generated at the factory during the calibration process can also be used to simplify the pressure calculation, essentially becoming a single linear equation that accounts for temperature changes in the sensor wires and enclosure. This mathematical implementation allows for easy switching to pressure calculations for other gas species by simply modifying the A thermal coefficient to match the other species being measured.

[0144] Furthermore, the thermal coefficient can be used to perform temperature compensation for both sensor lead and enclosure temperature variations. Reliance on a known thermal model allows improved temperature compensation to be performed over a wider temperature and pressure range. Such a thermal coefficient can also be used to improve the time response of a Pirani sensor. Conventional sensors generally rely on a temperature control loop to determine the sensor lead temperature; however, during events such as pressure steps or pulses, the sensor lead temperature may deviate from the target. Real-time and accurate measurement of the sensor lead temperature can detect such deviations from the target value and account for them in the pressure calculation.

[0145] The ability to update thermal coefficients originally collected and recorded at the factory while out in the field provides an excellent opportunity to detect changes in thermal characteristics that may indicate the end of useful life or the need for complete recalibration or repair. The vacuum gauge in the example embodiment adds the ability to vary the sensor wire temperature over a wide range, combined with the ability to T , T s , T e The measurement and thermal modeling of the T e The value changes can be quickly executed by T s Rapid changes in values, without any need for a dedicated vacuum system.

[0146] Zero offset adjustment

[0147] Pirani vacuum gauges are workhorse instruments used in the vacuum processing industry to measure gas pressures over a range that typically extends between 1E-4 Torr and 1E3 Torr. To ensure accuracy at the lowest pressures (i.e., P < 1E-2 Torr), it is necessary to perform daily "zero offset" (ZO) adjustments to compensate for drift due to contamination or corrosion of the sensor wires and changes in ambient or sensor wire temperature. The zero offset adjustment (ZOA) procedure is typically performed when the vacuum gauge is exposed to high vacuum levels (e.g., P ≤ 1E-6 Torr).

[0148] ZOA can be triggered in the field in several different ways. For example, a "zero" button may be used somewhere on the electronics module. The user pushes the button while in high vacuum and a new ZO correction value is stored in memory and used for subsequent measurements. Digital input pins can also be used to trigger ZOA. A logic pin can be tied to a PLC or triggered or ionized vacuum rail that controls the ZOA event. Alternatively, the ZO command may be available through a digital communications interface. In older vacuum gauges without microprocessor control, a potentiometer may be dialed until the pressure reading is correctly zeroed at high vacuum. Such a ZOA procedure generates a new ZO correction value that is stored in the controller's memory and then used to correct all subsequent pressure readings. For accurate operation, ZOA may be critical for all pressure readings P < 1E-2 Torr.

[0149] ZO drift can be minimized, but eliminating it may not be practical. Vacuum practitioners are aware of the need to frequently zero Pirani gauges. Therefore, a solution that reduces the frequency of ZOA would provide significant advantages. The following example implementation provides a new ZOA method based on a heat dissipation thermal model that provides several advantages for thermal conductivity vacuum gauges, including:

[0150] a) Lower minimum detectable pressure capability.

[0151] b) Improved accuracy when P<1E-2 Torr.

[0152] c) Ambient temperature specific correction of the ZO correction value (ie, reducing the ZO adjustment frequency).

[0153] d) Ability to update the zero offset correction temperature coefficient in the field (i.e., extend service life), referred to herein as "deep zero offset adjustment" (DZOA).

[0154] e) Ability to track changes in sensor wire characteristics due to corrosion or contamination (i.e., repair and predictive maintenance opportunities), thereby extending service life.

[0155] The physics-based causes of ZO drift are described below, followed by a discussion of how the above improvements are achieved in a thermal conductivity gauge that includes the following features:

[0156] a) Calibration and pressure measurement procedure based on thermal model.

[0157] b) Use three necessary and sufficient measurements: total power dissipation (W T ), sensor wire temperature (T s ) and the shell temperature (T e ).

[0158] c) Use temperature coefficients specifically designed for zero offset correction.

[0159] d) New ZOA options: Standard DOA (updates ZO correction value) and Deep ZOA (updates ZO correction value and its temperature coefficient).

[0160] As mentioned above, conventional vacuum gauges rely primarily on Wheatstone bridge circuit implementations and use the bridge voltage (V b ) as the critical pressure-dependent measurand that is calibrated to be pressure-dependent. In this case, the ZO adjustment simply measures the bridge voltage at high vacuum, stores the ZO correction voltage in memory, and subtracts the same value from all future bridge voltage measurements. Manufacturers often recommend frequent ZOA events, especially in more contaminated environments. ZOA is also required to measure accurate pressure at P < 1E-2 Torr if the ambient temperature changes by a few degrees.

[0161] Most conventional vacuum gauges have this ZOA capability but do not measure W as described above. T , T s , or T eThey also do not use calibration or pressure measurement methods based on heat sink thermal models. Conventional vacuum gauges also do not export or use temperature coefficients that are specifically calibrated and dedicated to ZO correction values. Some commercial vacuum gauges include some level of temperature correction or compensation of the pressure reading; however, those methods are limited to addressing self-heating issues at atmospheric pressure.

[0162] Back to Fig. 10A The "Total Losses" trace demonstrates an S-curve of total power consumption versus pressure over the typical operating range of a Pirani sensor. Heat dissipation in a Pirani sensor is governed by well-known thermal processes. Equation (12) above shows the total electrical power consumed by the sensor wire when the pressure and sensor wire temperature are stable. The first two terms W 端部 and W 辐射 is independent of pressure and corresponds to the end losses to the strut (W 端部 ) and radiation losses from the conductor to the surrounding (cooler) wall (W 辐射 ). These two terms can be referred to as “zero pressure power loss” and are mathematically expressed by equation 2 and T s and T e Related. 端部 corresponds to the end loss, and W 辐射 The term quantifies the radiation loss. Both terms depend on T through the following thermal coefficients s and T e : G and E as presented in equation (11) above. T (P=0) is the power offset, representing the power that the sensor wire would dissipate in the absence of gas molecules or gas heat conduction to the wall.

[0163] By definition, the total power loss at zero voltage, W T (P = 0), is independent of pressure - it provides a constant power offset to the total sensor lead power - and increases the power dissipated to the gas (W 气体 ). The total power loss at zero pressure is the fundamental reason why the ZOA function is required in all Pirani sensors.

[0164] exist Fig. 10A The “total loss” trace follows the total power dissipation, including all three terms of equation (12). As shown, the total power curve has a constant power offset: WT(P=0)≈400mW. As the gas pressure increases, the total power rises above the constant offset, clearly above 1E-3 Torr in the graph. The zero-pressure power of the vacuum gauge depends on several variables: (1) T e 、(2)T s , and (3) sensor wire condition, including size and surface finish.

[0165] Fig. 10A The "gas loss" trace in is obtained by subtracting the offset W from the total power due to thermal conductivity.T (P = 0) to isolate the power dissipated to the surrounding gas W 气体 . W at pressures below ≈1 Torr 气体 The linear behavior of W with respect to P response provides a simple way to calculate pressure (via linear equations) and perform temperature compensation. In fact, at low pressures, W 气体 It can be expressed as:

[0166] W 气体 = W T – W T (P = 0) = A * P * (T s (t)-T e ) / T e -1 / 2 (14)

[0167] Compared with previous methods, isolating W 气体 provides a convenient and simple calibration and pressure calculation procedure. However, this method relies on continuously tracking the W of the sensor. T (P=0). Power offset changes not captured by using ZOA may cause W 气体 An error occurred in the calculation and resulted in inaccurate pressure calculations.

[0168] It would be beneficial to minimize the effect of zero offset on the total power. Fig. 10A In the example shown, the power consumed by the gas (via thermal conductivity) matches the zero-pressure power offset at P ≈ 2E-2 Torr. This explains why measurements below 1E-2 Torr typically require an accurate and up-to-date W T (P=0) tracking to provide accurate W 气体 The value can then be used to calculate the low pressure. In other words, for pressures in the E-4 Torr range, W 气体 The contribution to total power dissipation is less than 1%, which means that an uncorrected drift of 1% in power offset will be reported as a pressure change on the order of E-4 Torr.

[0169] W T (P = 0) must be refreshed periodically using the ZOA procedure, exposing the vacuum gauge to high vacuum pressure and recording W T (P=0). However, not all vacuum processes are capable of using the high vacuum levels that may be required to provide consistent, accurate low pressure measurements.

[0170] According to equation (14), W T (P=0) is Ts and T e The thermal coefficients G and E can be determined at the factory as part of the sensor calibration procedure, thereby measuring their dependence on T s or T eBoth thermal coefficients can be derived from the mathematical fit as described above, stored in memory, and subsequently used to adjust the zero pressure power offset for temperature changes. Calculating and storing the independent thermal coefficients G and E at the factory as part of the calibration process is lacking in conventional methods and not only provides a means to perform temperature corrections for power offsets, but also provides a baseline measurement of sensor wire characteristics including: cleanliness and size.

[0171] Example embodiments may implement one or more ZOA processes. In the first and faster ZOA process, the vacuum gauge is exposed to high vacuum pressure conditions for a sufficient period of time to measure W T (P=0), T s and T e and store it in memory. This is a method that does not involve Ts and T e The thermal coefficients G and E are not updated, but the available records can be used to track W according to equation (14). T (P=0) with T e and T s changes.

[0172] In a second, longer process called "deep ZOA", the first ZOA process is performed as described above, and in addition, the ZOA is cycled through several T s The value is also recorded at a constant T e The G and E coefficients are updated by measuring (WT(P=0), Ts) at the same value. According to the functional form of equation (14), WT(P=0) is relative to T s (@Constant T e ) provides updated values ​​for the vacuum gauge's G and E. This update allows the user to achieve two goals:

[0173] a) Refreshes the temperature coefficients G and E to continue to provide accurate low pressure readings after changes in the bulk and surface characteristics of the sensor wire.

[0174] b) Track changes in G and E and report increases in PM that may warrant sensor replacement during the vacuum chamber's next preventive maintenance cycle. Conventional vacuum gauges do not have this predictive maintenance capability.

[0175] In the new (W(P=0),T s ,T e ) value set, and the updated G and E coefficients, the example implementation can continue to track the change of zero pressure offset with ambient temperature according to the above equation (14). Figure 7 The figure shows W T (P=0) relative to T s The curve fit of , produces the updated thermal coefficient value. Table 1 above shows the Figure 7 The coefficients derived from the nonlinear fit are shown by the trace in . Thus, the Deep ZOA process provides new G and E coefficients and the opportunity to diagnose end-of-life issues. Compared to the first ZOA process, which provides a new intercept for zero power, Deep ZOA provides both the intercept and slope of zero power with respect to temperature.

[0176] Changes in G indicate changes in thermal conductivity from the sensor wire to the column. This indicates changes in size and overall thermal conductivity due to etching, corrosion, or coating (i.e., contamination). Changes in E indicate changes in emissivity at the surface material of the sensor wire. This indicates changes in surface properties due to etching, corrosion, or coating. The vacuum gauge controller in the example embodiment can store acceptable boundaries for G and E values ​​for its sensors and issue service and maintenance reminders if the thermal coefficient approaches or exceeds these limits.

[0177] The above ZOA solution can provide several advantages. For example, the thermal coefficients G and E measured during factory calibration and stored in the vacuum gauge memory for temperature compensation of ZO provide better ZO temperature correction. The above first ZOA process improves on the previous method in that: 1) it collects and stores the three relevant measured quantities required to perform real-time temperature correction of ZO: W (P = 0), T s , T e , 2) Continue to temperature compensate the ZO using thermal coefficients G and E. Further, the "deep" ZOA process provides additional improvements that allow the G and E coefficients to be refreshed and account for changes in sensor wire temperature during operation. The updated coefficients effectively extend the life of the vacuum gauge, providing consistent ZO temperature compensation and accurate low pressure readings for longer periods of time. By comparing the G and E values ​​to the recommended limits of the sensor, the deep ZOA also provides the opportunity to issue repair and predictive maintenance alerts.

[0178] Vacuum gauge response time during fast pressure transients

[0179] Conventional Pirani gauge electronics rely on a Wheatstone bridge resistor network circuit (see Figure 1A) to operate a heated sensor wire (or heated component of a MEMS device) at a constant temperature. An analog feedback loop continuously adjusts the heating power the electronics direct to the sensor wire and balances the bridge while the sensor wire operates at a constant temperature (usually around 100°C). The bridge circuit adjusts the resistance of the sensor wire, which is factory calibrated to the operating temperature. The bridge voltage, V bis measured and used to calculate pressure based on a factory generated (or verified) calibration curve or lookup table. The ability of the balanced resistor bridge to control the sensor wire temperature is minimally challenged when the sensor is exposed to fairly constant gas pressure conditions. However, if sudden pressure transients (such as steps and pulses) occur, the wire temperature may briefly deviate from the nominal value. The magnitude and duration of the deviation depends on two factors: 1) the bandwidth of the temperature regulation feedback loop, and 2) the thermal capacity of the filament assembly. Prior art Pirani vacuum gauges do not monitor the wire sensor temperature, but rather assume that the temperature is essentially constant. Temperature deviations are not measured or detected in prior art vacuum gauges, and bridge voltage transients are always considered to be caused by pressure changes at an assumed constant filament temperature.

[0180] The vacuum gauge in the example embodiment can improve the ability to accurately track fast pressure transients in real time, thereby bringing benefits to vacuum processing applications. A common pressure transient event in a vacuum chamber is the rapid exhaust process to atmospheric pressure that occurs in a load lock. During the rapid exhaust, the influx of gas into the sensor chamber suddenly increases the thermal conductivity from the heated wire to the adjacent wall and drives the wire temperature to drop suddenly, which is undetectable by any traditional vacuum gauge. The feedback loop responds by delivering additional heating power for two independent purposes: 1) returning the filament to its nominal temperature (i.e., an effect independent of gas pressure), and 2) adjusting the heating power delivery to meet the increased thermal conductivity demand (i.e., an effect related to gas pressure). Conventional vacuum gauges do not detect wire temperature changes that occur during transients, but assume that the filament temperature is unchanged and the bridge voltage transient is caused only by gas thermal conductivity / pressure changes. In other words, the pressure report during the transient is inaccurate and delayed due to the need to restore the filament temperature. The response time associated with temperature recovery is related to the power delivery capability of the sensor, the bandwidth of the feedback loop, and the thermal capacity of the sensor assembly. Pirani sensors with smaller sensor heat capacity specifications perform better in tracking fast transients. Minimizing heat capacity typically drives sensor wire circuit design and explains the enhanced dynamic response of MEMS sensors. MEMS sensors have the smallest Cs values ​​in the industry and are cited by applications that experience fast pressure transients. Example implementations are compatible with all brands of thermal conductivity sensors (including wire designs and MEMS scale designs). The process described herein can also be applied to sensors where the controller maintains a heated component at a constant temperature or where the temperature of the heated component depends on the pressure. The model works in all those cases.

[0181] An example embodiment provides a method for measuring pressure using a thermal conductivity vacuum gauge based on a thermal model. The temperature of the sensor wire and the enclosure can be measured at any time, and the electrical power delivered to the sensor wire can also be measured, while an additional thermal coefficient representing the heat capacity of the sensor wire is included in the pressure calculation algorithm to isolate any power directed to restore the sensor wire temperature during pressure transients. The effect of including the heat capacity term in the pressure calculation improves the dynamic response of the Pirani sensor during the measurement of fast gas pressure events. The amount of power directed to restore the sensor wire temperature is proportional to the sensor wire heat capacity and the rate of change of the wire temperature. Rapid wire temperature measurement is beneficial in providing a more accurate measurement of the heating power driven by the heat capacity. Several methods of measuring the heat capacity coefficient of the sensor wire are also described. Such methods can be used during the manufacturing process. Routine measurement of heat capacity on production units provides an additional way to identify incoming materials and filter out unsuitable materials from the production line.

[0182] The above describes the method of measuring total power (W) by combining thermal model calibration and calculated pressure. T ), sensor wire temperature (T s ) and the shell temperature (T e Next, this solution is extended by including an additional term in the total power equation that accounts for the heating power that is directed into the sensor to readjust the sensor temperature during pressure transients.

[0183] Heat capacity correction

[0184] The heat capacity of the sensor lead assembly plays a key role in the dynamic response of the Pirani sensor to pressure transients. The heat capacity of the sensor lead assembly is defined as:

[0185] C s = Q T / ΔT (15)

[0186] in,

[0187] C s = Heat capacity of sensor lead assembly

[0188] (For example, the specific heat capacity of tungsten is 132 J / K kg)

[0189] Q T = quantity T s The amount of energy required to change ΔT.

[0190] The total power consumed by the sensor lead assembly during operation takes into account not only the power due to gas, radiation, and terminal losses, but also the power required to adjust T sIn fact, if T s (t) is not a constant, even if the gas pressure is fixed, the total power W T (t) is also not a constant, and so the power delivered to correct for the sensor wire temperature must be incorporated into the total power equation:

[0191] W T (t) = W 端部 + W 辐射 + W 气体 + C s [dT s / dt] (16)

[0192] W T (t) = G*(T s (t)-T s )+E(T s 4 (t)-T e 4 )+A*P*(T s (t)-T e )T e -1 / 2 )+C s *

[0193] [dT s / dt](17)

[0194] In the above equation, C s [dT s / dt] is the additional power term that isolates the heating power consumed by the change in conductor temperature. s The amount of energy required is: dQ = Cs*dT as expressed in equation (15) s .

[0195] The heat capacity of the sensor wire assembly is an important thermal coefficient when considering the dynamic response of a thermal conductivity vacuum gauge. The heat capacity includes the effect of the specific heat capacity of the sensor wire and some additional effects of the attachment post. During a pressure transient (e.g., a pressure burst), the sensor wire typically cools briefly and the feedback loop responds by applying additional heating power to restore its desired temperature. Unless the additional power associated with the sensor wire temperature change is isolated and accounted for in the power equation, this will cause errors in the pressure calculation. Conventional Pirani vacuum gauges do not track T s, so a conventional Pirani vacuum cannot distinguish between gas pressure transients and the resulting wire temperature changes. Heat capacity can be particularly important in the model-based pressure calculations described in this article if the sensor wire is not operated at a constant temperature, or if the heat capacity is very large or the bandwidth of the heating circuit is very small.

[0196] Fig.12 is a plot of normalized pressure versus time and illustrates the improvement in dynamic response that can be achieved in an example vacuum gauge once the heat capacity coefficient is incorporated into the pressure calculation above. The diamond trace represents the actual pressure transient reported by the capacitance pressure gauge. The square trace represents the pressure transient measured without considering the last term in equation (17). The circular trace shows the reported pressure once the full equation is considered as in equation (17). This gain in time response is particularly beneficial for sensors with large heat capacities or low bandwidth heater circuits. The ability to measure fast transients with large heat capacity sensors allows for the design of more robust vacuum gauges, including additional materials (e.g., thicker wire) that can withstand process chemistry and provide longer service life.

[0197] C s Measurement - Cooling time constant

[0198] Heat capacity C s It is important to calculate the time required for the sensor wires to cool when power is removed. The time required for the sensor wires to cool after power is removed is given by the following equation:

[0199] T s (t) = T e + (T s, nom -T e ) * e -(a *t / Cs) (18)

[0200] in:

[0201] T s (t): Sensor wire temperature as a function of time.

[0202] T e : Final temperature (cold wire)

[0203] T s,nom : Initial and nominal wire temperature (hot wire)

[0204] C s : Heat capacity of sensor wire assembly

[0205] a: Thermal conductivity: G+A*P*T e -1 / 2

[0206] If the heat dissipation (e.g., end losses and gas thermal conductivity) increases or the heat capacity decreases, the cooling time of the sensor wire will be faster. In fact, the time constant of the cooling process is given by:

[0207] T(P) = C s / a = Cs / (G + A *P* T e -1 / 2 ) (19)

[0208] Equation (19) shows the cooling time constant as a function of gas pressure, with the longest time constant occurring at zero pressure.

[0209] C s Measurement - Pressure Rise

[0210] C s It can be used to determine the time it takes for the sensor wire assembly to reach a desired temperature for a given heating power delivery step. As indicated above, the temperature rises exponentially at constant power, and the time constant increases with pressure because less power is available to heat the sensor wire.

[0211] A first approximation of the heat capacity of the wire can be obtained using the dimensions and specific heat capacity of the W material (e.g., length 2 inches, diameter 0.0005 inches, specific heat capacity: 132 J / Kg K, density: 19300 Kg / m 3 : Cs = 1.64E-5J / K). This number is not expected to be an adequate representation of the heat capacity of the entire sensor assembly, as the support is expected to contribute to the overall heat capacity. Using this calculated Cs number and the average G = 5E-6W / K under zero pressure conditions, this shows that the time constant is:

[0212] T(P=0)=1.64E-5 / 5E-6≈3.3 seconds.

[0213] This indicates that the filament will remain hot for many seconds after power is removed under zero pressure conditions, and that the time will decrease in proportion to the increase in pressure. Conversely, the higher the gas pressure, the longer it will take to heat the filament.

[0214] C s Measurement: Calorimetric determination

[0215] A straightforward way to perform heat capacity measurements is to perform a calorimetric measurement at zero pressure. In a calorimetric experiment, an electrical heating power step is applied to the sensor wire assembly while tracking its temperature increase over time: T ref (t). For each measurement, the power dissipated to thermal conductivity (i.e., minus the end losses at zero pressure) is subtracted from the total power. The remaining power (used to heat the sensor wire assembly) is then integrated and divided by the temperature rise it causes to provide a direct measure of the sensor wire heat capacity: Cs . At zero pressure:

[0216]

[0217] If the power W T During the heating process, the temperature of the sensor wire assembly increases according to the following equation when stepped and controlled to a constant value:

[0218] T s (t) = T e + (W T / a) * (1 – e (-t / Τ) ) (twenty one)

[0219] in:

[0220] Τ=C s / a, time constant of temperature rise

[0221] a=thermal conductivity: G+A*P*T e -1 / 2

[0222] P = Gas pressure

[0223] Fig.13 is a graph of sensor wire temperature versus time and shows the results of a calorimetric experiment where, with a current (power) step, T s From 90°C to 100°C. The sensor starts at 90°C and the heating current suddenly steps to the value corresponding to T s = 100°C. Since the resistance of the sensor wire does not change significantly over this temperature range, the power delivery is assumed to be constant and the data can be fit to equation (21). This fit shows the following: 1) T = C s / G = 5.468 seconds, 2) G = 5.2E-6 W / K, and 3) Cs = 2.85E-5 J / K. Here, the measured value exceeds the calculated value: 1.64E-5 J / K confirming that the heating above ambient temperature extends beyond the sensor wire, i.e., into the sensor wire assembly. For this reason, C s Can be referred to as the thermal capacity of the sensor wire assembly.

[0224] Fig.14 is a graph of the sensor temperature versus time and shows an additional calorimetric experiment where Ts was increased from 20°C to 100°C. The time constant is 5.49 seconds (C s =2.86E-5), which is very consistent with the data discussed above.

[0225] FIG. 15A to FIG. 15Dis a graph of sensor temperature versus time, showing the typical sensor operating temperature between 90°C and 110°C (i.e., nominal operating T ref A series of time constants for the gas pressure between zero pressure and 0.1 Torr are shown in the figure. As shown, the time constant decreases with increasing pressure because additional power is required to maintain the wire at the required temperature. In all cases, C is calculated using the known values ​​of s : G, A, P and T e , using the following relationship:

[0226] Τ = C s / (G + A*P * T e -1 / 2 ) (twenty two)

[0227] in,

[0228] G=5.2E-6W / K

[0229] A=0.00366

[0230] Table 3 below shows the calculated Cs values ​​at different pressures. The consistency of the Cs values ​​indicates that equation (22) is the correct physical model for the time constant.

[0231] Pressure Time constant (Τ), seconds Cs[J / K] Zero pressure 5.46919 2.86E-5 1E-3 5.37709 2.91E-5 1E-2 3.8532 2.79E-5 1E-1 0.969973 2.49E-5

[0232] Table 3

[0233] Fig.16 is a plot of (1 / T) versus P, which should yield a linear relationship according to the following equation (23):

[0234] (1 / Τ) = (G / Cs) + (A*T e -1 / 2 / Cs) * P (23)

[0235] in:

[0236] Intercept = G / Cs, Eq. 20 → Cs = 5.2E-6 / 0.1742 = 2.98E-5 J / K

[0237] Slope = A*T e 1 / 2 / Cs, Eq. 21->A=0.00437

[0238] When the value of G is known, C can be calculated using the intercept of 1 / T with respect to P. s And the slope can be used to export A.

[0239] Fig.17 is a graph of heat capacity as a function of wire temperature. Heat capacity Cs It can be further used to s Characterization of the temperature dependence of: Cs(T s In one example, using a 10°C sensor wire temperature step around the target temperature (ie, + / - 5°C), the thermal capacity can be calculated at several different values ​​of Ts around the nominal operating setting. Fig.17 Established the relationship between heat capacity and T s Linear dependence of:

[0240] C s (T s )=5.1E-8[J / K 2 ]*T s [K]+9.629E-6J / K

[0241] C s (T s =373K)=2.87E-5J / K

[0242] The dependence of Cs on the sensor wire temperature is likely related to the temperature gradient of the wire with T s As the wire temperature increases, a larger portion of the assembly is bonded and the thermal capacity of the extended thermal system increases.

[0243] Overview of the process for determining heat capacity

[0244] Given the above, an example procedure for determining the thermal capacity of a sensor wire is as follows:

[0245] a) Determine α: Measure the wire resistance versus temperature in the absence of self-heating.

[0246] b) Determine G and E: According to T ref , measure power W at zero pressure T →Determine ε based on E.

[0247] c) Measure W based on power in the middle of the low pressure range and in the linear pressure range T (P) Determine A.

[0248] d) Determine C from the exponential time constant for heating the wire at zero pressure to two reference temperatures: s :T 低 →T 高 .

[0249] The power delivered to the sensor is expressed according to equations (16) and (17) above. For a brand new vacuum gauge and a narrow temperature range around ambient, the contribution of radiation losses is 10%, and equation (17) can be further simplified:

[0250] WT (t) = G o *(T s (t)-T e )+A*P*(T s (t)-T e )T e -1 / 2 )+C s *[dT s / dt] (twenty four)

[0252] Here, G o = linear temperature coefficient of zero-pressure power loss, including radiation losses over the narrow temperature range typical for the operating range of Pirani sensors (ie, 0-60°C).

[0253] Power term C s *[dT s / dt] provides an efficient way to subtract the gas independent heating power (consumed in adjusting the sensor wire temperature) from the total power consumed by the sensor. The remaining power can then be used to accurately calculate the gas pressure during rapid pressure events (such as overpressure or atmospheric exhaust events). As mentioned above, since the heater can only provide source power, equation (17) applies, and [dT s / dt]>0.

[0254] Incorporating the heat capacity thermal coefficient into the design, calibration, and pressure calculations of Pirani sensors provides several unique advantages as described below.

[0255] Fast Pirani sensors are typically designed to have small heated components, especially on a MEMS scale. For wired designs, a small wire diameter is generally preferred because this minimizes end losses due to thermal conductivity and improves response time due to reduced heat capacity. From a thermal modeling perspective, the reduction in size addresses the fact that small sensors have smaller sensor wire component heat capacity and recover more quickly from sensor wire temperature changes, requiring less thermal energy to restore their nominal sensor wire temperature. By additionally measuring the sensor wire temperature, and taking into account the heat capacity term (in equation (17)), example embodiments can provide fast time responses to fast events such as atmospheric exhaust, even with larger sensor wire configurations.

[0256] Fig.18is a graph of normalized pressure versus time and demonstrates the performance of a MEMS-scale prior art sensor compared to the heat capacity compensated sensor in the example embodiment. The diamond trace is the actual pressure transient measured by a capacitance pressure gauge with a response time of 10 milliseconds. The triangle trace is the pressure response measured using a fast MEMS-scale Pirani vacuum gauge. The circular trace is the response of a standard gold-coated tungsten wire Pirani, including 1) real-time sensor wire temperature measurement and 2) heat capacity compensation as indicated by the last term of Equation 13. The square trace represents a wire Pirani vacuum gauge with a fast temperature control system. As shown in the figure, the operating goal of this sensor is to eventually catch up with the pressure transient, but it will lag (i.e., underestimate the pressure) until the heater circuit begins to operate to restore the wire temperature. The inclusion of sensor wire temperature measurement and the heat capacity term for pressure calibration allows the vacuum gauge with larger sensor wires in the example embodiment to match the dynamic response of one of the fastest conventional Pirani sensors.

[0257] In particular, larger sensor wires naturally contain more metal material, are longer and have a larger diameter. A larger diameter means a more robust mechanical design and better compatibility with vibrations common in vacuum systems. A large diameter also means increased robustness against reactive chemical products that may attack metal surfaces, providing longer sensor life in the presence of corrosive species.

[0258] Example embodiments also provide advantages in terms of calibration. The thermal capacity C is performed during the calibration process. s The measurement of not only provides the thermal coefficient needed to perform a more accurate pressure calculation via Equation (17), but also provides another measure of the overall thermal characteristics associated with the sensor wire construction. Heat capacity measurements can be used to track the manufacturing process, looking for changes in the thermal properties of the sensor wire. Sudden changes or long-term drift in the sensor's heat capacity indicate a manufacturing problem, such as changes in wire size or material composition.

[0259] Example embodiments also provide the benefit of a vacuum gauge for measurement. Incorporating the heat capacity of the sensor into the pressure calculation, equation (17) combined with real-time sensor wire temperature monitoring provides faster pressure measurement response and accuracy in case of pressure transients. Pirani sensors are used daily in load lock stations where rapid venting is the norm and rapid response is critical to improving the speed and accuracy of those events.

[0260] While example embodiments have been particularly shown and described, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the embodiments as encompassed by the appended claims.

Claims

1. A thermal conductivity vacuum gauge, include: an envelope enclosing a gas volume; a sensor wire positioned within the gas volume; as well as A controller configured to: providing a power dissipation model for the thermal conductivity vacuum gauge, the power dissipation model including power losses due to heat conduction losses at sensor wire end contacts, radiation losses from the sensor wire to the gas enclosure, and pressure-dependent heat conduction losses from the sensor wire through the surrounding gas; applying a power input to the sensor wire to heat the sensor wire; Measure the total power dissipated during the application of this power input, W T , sensor wire temperature T s , and the shell temperature T e ;as well as Based on the measured W T , T s , and T e and the power dissipation model determine the gas pressure within the enclosure.

2. The thermal conductivity vacuum gauge according to claim 1, in, The thermal conductivity vacuum gauge is a Pirani vacuum gauge.

3. The thermal conductivity vacuum gauge according to claim 1, in, The controller is further configured to: At a subsequent time, the total power dissipated during the application of this power input is measured, W T , sensor wire temperature T s , and the shell temperature T e and A change over time of at least one of the end loss coefficient G and the radiation loss coefficient E is determined based on the subsequent values.

4. The thermal conductivity vacuum gauge according to claim 1, in, The controller is further configured to: Measure the total power dissipation W when the power input is different T and the sensor wire temperature T s Multiple different values ​​of ; Determine the total power dissipation W T and the sensor wire temperature T s A mathematical fit of the multiple different values ​​of ; and The values ​​of the end loss coefficient G and the radiation loss coefficient E are determined based on this mathematical fit.

5. The thermal conductivity vacuum gauge according to claim 4, in, When the power input is different, the shell temperature T e Keep at a constant value.

6. The thermal conductivity vacuum gauge according to claim 4, in, The controller is further configured to output a notification to remove and replace the sensor wire based on a comparison of at least one of the end loss coefficient G and the radiation loss coefficient E with a reference value.

7. The thermal conductivity vacuum gauge according to claim 4, in, The controller is further configured to: determining a gas adaptation factor A based on the mathematical fit, the gas adaptation factor A depending on the type of gas within the enclosure; and Based on the gas adaptation factor A, a measure of the gas pressure within the enclosure is determined.

8. The thermal conductivity vacuum gauge according to claim 1, in, The controller is further configured to: Based on the sensor wire temperature T s The change in a given period of time determines the sensor wire C S The heat capacity of Based on the sensor wire C S The heat capacity determines the measure of the gas pressure within the enclosure.

9. The thermal conductivity vacuum gauge according to claim 8, in, The controller determines the sensor wire C based on the cooling rate of the sensor wire temperature Ts within the given time period. S heat capacity.

10. The thermal conductivity vacuum gauge according to claim 8, in, The controller determines the gas pressure measure during an increase in gas pressure within the housing.

11. A method of operating a thermal conductivity vacuum gauge comprising a sensor wire in a gas volume within an enclosure, the method include: providing a power dissipation model for the thermal conductivity vacuum gauge, the power dissipation model including power losses due to heat conduction losses at sensor wire end contacts, radiation losses from the sensor wire to the gas enclosure, and pressure-dependent heat conduction losses from the sensor wire through the surrounding gas; applying a power input to the sensor wire; Measure the total power dissipated during the application of this power input, W T , sensor wire temperature T s , and the shell temperature T e ;as well as Based on the measured W T , T s , and T e and the power dissipation model determine the gas pressure within the enclosure.

12. The method according to claim 11, in, The thermal conductivity vacuum gauge is a Pirani vacuum gauge.

13. The method of claim 11, further comprising: include: At a subsequent time, the total power dissipated during the application of this power input is measured, W T , sensor wire temperature T s , and the shell temperature T e and A change over time of at least one of the end loss coefficient G and the radiation loss coefficient E is determined based on the subsequent values.

14. The method of claim 11, further comprising: include: Measure the total power dissipation W when the power input is different T and the sensor wire temperature T s Multiple different values ​​of ; Determine the total power dissipation W T and the sensor wire temperature T s A mathematical fit of the multiple different values ​​of ; as well as Based on the mathematical fit, values ​​of the end loss coefficient G and the radiation loss coefficient E are determined, and the power dissipation model combines the end loss coefficient G and the radiation loss coefficient E.

15. The method of claim 14, in, When the power input changes, the shell temperature T e Keep at a constant value.

16. The method of claim 14, further comprising selectively removing and replacing the sensor wire based on a comparison of at least one of the tip loss coefficient G and the radiation loss coefficient E with a reference value.

17. The method of claim 14, further comprising: include: Determining a gas adaptation factor A based on the mathematical fit, the gas adaptation factor A depending on the type of gas in the housing; as well as Based on the gas adaptation factor A, a measure of the gas pressure within the enclosure is determined.

18. The method of claim 11, further comprising: include: Based on the sensor wire temperature T s The change in a given period of time determines the sensor wire C S The heat capacity of Based on the sensor wire C S The heat capacity determines the measure of the gas pressure within the enclosure.

19. The method of claim 18, in, Based on the sensor wire temperature T s The cooling rate during this given period of time determines the sensor wire C S heat capacity.

20. The method of claim 18, in, A measure of the gas pressure during the increase in gas pressure within the housing is determined.

21. The method of claim 11, further comprising modeling the zero offset of the thermal conductivity vacuum gauge by: evacuating the enclosure to substantially zero pressure; applying a power input to the sensor wire; Measure the total power dissipated during the application of this power input, W T , sensor wire temperature T s , and the shell temperature T e ; as well as According to W T , T s and T e Determine the end loss coefficient G and radiation loss coefficient E in the power dissipation model.

22. The method of claim 21, in, The end loss factor G corresponds to the heat losses at the end posts to which the sensor wire is coupled during application of the power input to the sensor wire.

23. The method of claim 21, in, The radiation loss coefficient E corresponds to the radiation loss of the sensor wire during application of the power input to the sensor wire.

24. The method of claim 21, further comprising: include: Determining a temperature compensation value that compensates for changes in at least one of a sensor wire temperature and a package temperature, the temperature compensation value being a function of at least one of the tip loss coefficient G and the radiation loss coefficient E; as well as The gas pressure measure is determined based on the temperature compensation value.

25. A method of operating a thermal conductivity vacuum gauge, include: applying a power input to the sensor lead when the sensor lead and the housing of the enclosure housing the thermal conductivity vacuum gauge exhibit substantially zero pressure; Measure the total power dissipated during the application of this power input, W T , sensor wire temperature T s , and the shell temperature T e ; Based on the sensor wire temperature T s The change in a given period of time determines the sensor wire C S The heat capacity of as well as Based on the heat capacity C of the sensor wire S And a measure of the power input applied to the sensor wire determines a measure of the gas pressure within the housing.

26. The method of claim 25, in, The sensor wire C is determined based on the cooling rate of the sensor wire temperature Ts within the given time period. S heat capacity.

27. The method of claim 25, in, A measure of the gas pressure during the increase in gas pressure within the housing is determined.

28. A thermal conductivity vacuum gauge, include: an envelope enclosing a gas volume; a sensor wire positioned within the gas volume; as well as A controller configured to: applying a power input to the sensor lead when the sensor lead and the housing of the enclosure housing the thermal conductivity vacuum gauge exhibit substantially zero pressure; The total power dissipated during the application of the power input that heats the sensor wire is measured as W T , sensor wire temperature T s , and the shell temperature T e ; Based on the sensor wire temperature T s The change in a given period of time determines the sensor wire C S The heat capacity of as well as Based on the heat capacity C of the sensor wire S And a measure of the power input applied to the sensor wire determines a measure of the gas pressure within the housing.

29. The thermal conductivity vacuum gauge according to claim 28, in, The controller is based on the sensor wire temperature T s The cooling rate during this given period of time determines the sensor wire C S heat capacity.

30. The thermal conductivity vacuum gauge according to claim 28, in, The controller determines the gas pressure measure during an increase in gas pressure within the housing.

31. A method of operating a thermal conductivity vacuum gauge comprising a sensor wire in a gas volume within an enclosure, the method include: The zero offset of this thermal conductivity vacuum gauge is modeled as follows: evacuating the enclosure to substantially zero pressure; applying a power input to the sensor wire; Measure the total power dissipated during the application of this power input, W T , sensor wire temperature T s , and the shell temperature T e ; as well as According to W T , T s and T e Determine the end loss coefficient G and radiation loss coefficient E; Providing a power dissipation model of the thermal conductivity vacuum gauge, the power dissipation model combining the end loss coefficient G and the radiation loss coefficient E; applying a power input to the sensor wire; Measure the total power dissipated during the application of this power input, W T , sensor wire temperature T s , and the shell temperature T e ;as well as Based on the measured W T , T s , and T e and the power dissipation model determine the gas pressure within the enclosure.

32. The method of claim 31, in, The power dissipation model further includes power losses due to conductive heat losses of the sensor wire end contacts, radiation losses from the sensor wire to the gas enclosure, and pressure-dependent conductive heat losses from the sensor wire through the surrounding gas.

33. The method of claim 31, in, The end loss factor G corresponds to the heat losses at the end posts to which the sensor wire is coupled during application of the power input to the sensor wire.

34. The method of claim 31, in, The radiation loss coefficient E corresponds to the radiation loss of the sensor wire during application of the power input to the sensor wire.

35. The method of claim 31, further comprising include: Determining a temperature compensation value that compensates for changes in at least one of a sensor wire temperature and a package temperature, the temperature compensation value being a function of at least one of the tip loss coefficient G and the radiation loss coefficient E; as well as The gas pressure measure is determined based on the temperature compensation value.

Images