Hoffman D.M., Singh B., Thomas J.H. (Eds). Handbook of Vacuum Science and Technology

Подождите немного. Документ загружается.

380

Chapter 3.4: Mass Flow Measurement and Control

tube in laminar flow is related to tht fourth power of diameter. For the level of

flow accuracy desired for these instruments, better than 1%, the mechanical toler-

ances are effectively unachievable unless calibration is performed after assembly

of the individual sensor and bypass combination.

3.4.2.3

Control Valve Design

The sensor, associated electronics, and bypass (where one is needed) form a mass

flow meter. To make the instrument a mass

flow

controller,

control valve is placed

downstream from the meter and a control circuit is added to the electronics. Con-

trol valves used in typical MFCs allow continuous adjustment of the conductance

as opposed to the on/off action of a shutoff valve. The resolution of conductance

is much better than 0.1% of the open conductance to obtain the desired control

range of the MFC. Generally, the control valve is not capable of providing posi-

tive

shutoff,

and a separate valve intended for that purpose should be used. The

valve's open conductance can be set by selecting the orifice diameter and, as with

the bypass, is determined by the full-scale range of the MFC. Two varieties of

control valves are in common

use:

solenoid actuated and piezoelectric actuated.

Solenoid valve designs are more traditional and more commonly used in MFCs.

An overview of the typical construction of a solenoid valve is shown in Figure 3.

The operating principle relies on current in the solenoid to generate a magnetic

field that acts on the armature to open the valve. A flat spring is normally under

Fig.

Coil

Magnetic

Armature

Orifice

Solenoid valve overview.

3.4.2 Overview of Thermal Mass Flow Controller Technology

381

Fig.

Orifice

Piezo valve overview.

tension (preload) to oppose the force provided by the solenoid actuator. This

configuration results in a normally closed valve whereby if power is removed the

device will assume a closed condition. Normally open valves can also be designed

with solenoid actuation, but they tend to be more complicated and are not nearly

as common.

Flow controllers with piezo valves were developed in the mid-1980s and have

been used primarily in applications where high purity and corrosion resistance

are primary concerns. The construction features of the piezo valve are repre-

sented in Figure 4. The piezo actuator is a stack of 50-100 individual disks made

from a special ceramic material that exhibits the piezoelectric effect where an elec-

tric field generated across it causes the material to expand along that axis. Inter-

digitated electrodes generate the electric field when voltage is applied to the actu-

ator terminals. This causes each disk to expand, and the combined effect acts

to close the valve. Thus, piezo valves tend to be designed to operate as normally

open, although normally closed versions are also possible. Also, their capability

for generating high force allows piezo actuators to be used in diaphragm isolation

valve designs as shown in Figure 4.

3.4.2.4

Control System Design

The control circuit is based on PID (proportional integral derivative) feedback on

the controllor error signal generated by the flow sensor versus the set point signal

382 Chapter 3.4: Mass Flow Measurement and Control

supplied to the MFC. Note that the controUor error is simply the instantaneous

difference between the flow signal and the set point; it should not be confused

with the overall instrument error, which is represented by the difference between

the flow signal and the actual flow. The control action is tuned to drive the valve

so that the controller error is reduced to zero as rapidly as possible with little or

no overshoot. Tuning involves changing parameters governing the action of the

PID gains to achieve the desired response to set point changes. The tuning can

be optimized for a given set of conditions such as gas type, upstream pressure,

downstream pressure, physical orientation, and set point; but if the conditions

change significantly, controller performance may be affected. Symptoms are typ-

ically slow response to set point changes or control oscillation.

3.4.3

PERFORMANCE CHARACTERISTICS

The proper installation procedure, gas line components, and ancillary electronic

instrumentation help ensure optimal performance and reliability from an MFC.

Environmental conditions such as temperature, orientation, vibration, and electro-

magnetic interference may also affect an MFC's ability

measure and control flow.

3.4.3.1

Installation

For all gas line components including the MFC, the proper fittings and mechani-

cal installation are necessary to provide a leak-tight system. The fittings on most

MFCs are either face seal (VCR®, registered TM of Cajon Co., Macedonia, OH)

or ferrule seal (Swagelock®, registered TM of Crawford Fitting Co., Solon, OH)

for

j-inch

pipe. However, if the full-scale flow is greater than 20 slm, the fittings

may be for ^-inch pipe. For vacuum system or high-purity applications, the MFC

fittings should be helium leak tested if possible. This is extremely important in

the case of toxic or corrosive gas service. In addition to ensuring the leak integrity

of the system, proper purging to reduce atmospheric contamination (moisture and

oxygen) to acceptable trace levels is crucial for reliable performance. The specific

method to achieve this condition depends on the system and level of contamina-

tion desired. At a particular facility, standard purging procedures should be estab-

lished for each case. Likewise, procedures for removing the toxic or corrosive gas

before removing an MFC (or any other gas line component) must be established

and followed.

3.4.3 Performance Characteristics

383

After the recommended warmup period, a check of the indicated zero is advis-

able.

The user's manual for the particular instrument should be consulted regard-

ing the proper procedure, but in general, the following guidelines may be used:

• Ensure there is no flow by closing a separate shutoff valve in line with the

MFC.

• Command a set point flow to the MFC. (This equalizes the pressure upstream

to downstream.)

• After the indicated flow has stabilized, command a set point flow of zero.

• Adjust the potentiometer, usually accessible from the outside of the instru-

ment, labeled "zero."

• Alternatively, the adjustment may be performed at the readout/set point

device.

This process avoids the problem of "zeroing in" a small leak, which may nor-

mally occur through the MFC control valve while a differential pressure remains

across the MFC during the zero adjustment.

3.4.3.2 Gas Line Components

In line with the MFC is often a combination of other instruments and components

including a pressure regulator, a particle filter, shutoff

valves,

and a pressure trans-

ducer. In some industries, such as semiconductor manufacturing, this is called a

gas stick. A typical gas stick layout is shown in Figure 5. The pressure regulator

and filter can affect MFC performance if a fluctuation in pressure occurs or if the

differential pressure becomes too low or too high.

3.4.3.3 Electrical Requirements

The electrical connector on an MFC is usually either a "card edge," 15-pin D, or

9-pin D. The D connectors provide a degree of electromagnetic shielding when

Fig.

Inlet

Pressure

Regulator

Shutoff Filter

Valve

Shutoff

Valve

Outlet

Typical gas stick.

384 Chapter 3.4: Mass Flow Measurement and Control

used with a shielded cable. Unfortunately, there is no standard for the electrical

connections (pinouts) between manufacturers, so caution must be used to ensure

that proper electrical compatibility is maintained among power supply, cable, and

MFC.

The industry standard power supply for MFCs is +/—

V DC with a typi-

cal current draw of 100 mA (200 mA maximum at startup). The set point voltage

and indicated flow are each 0 to 5 V, but it should be noted that most instruments

are capable of indicating some amount of reverse flow (negative voltage) and con-

trolling to well above full-scale flow (greater than 5 V).

Other functions that may be available as signals at the connector include com-

mands for overriding the internal valve control to drive it completely open or

closed; a voltage for monitoring the internal valve drive signal (an important di-

agnostic); an optional control signal input; and a 5 V reference.

There are usually three separate common lines with separate functions. The

signal common is used as a 0 V reference for the set point and indicated flow. The

power common is used to return current imbalances in the power supply draw. A

chassis ground is intended to provide a connection to the body and case of the in-

strument. Although some designs are more tolerant of combining the signal com-

mon and power common, it is recommended to keep these three connections sep-

arate in any precision instrumentation such as an MFC.

3.4.3.4 Temperature Effects

The ways in which the environmental temperature may affect the behavior or per-

formance of an MFC include shifts in the indicated zero, shifts in the indicated

span or sensitivity, reduced performance of the control capability, or inability to

reach set point due to valve conductance loss.

The most easily recognized problem due to temperature effects is a zero shift

expressed quantitatively as a zero temperature coefficient (zero TC). The specifi-

cation for zero TC is usually less than 0.05 % FS/°C. So for example, if the tem-

perature around an MFC changed by 10°C (18°F) the zero could change by as

much as 0.5 % FS. This can be compensated for by re-zeroing the device at the

operating temperature following the guidelines in Section

3.4.3.1.

The shift in span or sensitivity as a function of temperature is more difficult to

recognize and account for in a typical application. For gases that do not exhibit a

large temperature dependence of

(see Section 3.4.1.2) the span TC is typically

less than

0.1%

of Rdg/°C. However, gases having a temperature sensitive c^ will

exhibit additional span TC behavior.

The operating environmental temperature range for MFCs is usually rated at

0 to 50°C.

3.4.3 Performance Characteristics

385

3A3.5 Orientation (attitude) Effects

Since flow measurement and control instrumentation must be mounted in a vari-

ety of locations and orientations relative to gravity, the effect known as "thermal

siphoning" may be exhibited in MFCs. This effect occurs when the heated sensor

is oriented other than horizontally and the buoyancy difference between the gas in

the sensor and bypass causes a circulating flow within the MFC. Although the ef-

fect is present anytime such orientation exists, it is most noticeable when a par-

ticularly dense gas is used (e.g., SF6, Kr, or the various MFCs and CFCs and re-

lated compounds). The effect is also line pressure sensitive, with higher pressures

resulting in higher circulating flows.

For small offsets, the thermal siphoning influence exists independent of the

flow through the device and therefore acts as a fixed offset to all flow measure-

ments. In this case (less than 5% FS shift), the device may be re-zeroed at line

pressure by following the guidelines in Section 3.4.3.1 and used normally.

3.4.3.6 Gas Type Effects

The type of gas flowing through an MFC has a direct effect on the calibration of

the measurement and the control performance.

A gas correction factor (GCF) may be applied to the output signal to convert

the sensitivity from the calibration gas to another gas. Recommended GCF values

are listed in most user's manuals and may vary from one manufacturer to another.

Caution must be exercised when using GCFs as the accuracy of the standard gas

correction method is not better than \% for many gases and for certain situa-

tions may introduce as much as 10% error. The accuracy of a particular gas cor-

rection depends on several factors, including the two gases involved, the instru-

ment model, and the range (bypass or no bypass). For a particular case, specific

recommendations may be available from the instrument manufacturer.

Since the issue of gas correction is, practically speaking, very complex, an

ideal approach is to perform a calibration or verification with the actual gas in the

actual environment if

possible.

For some applications, in situ or point-of-use cali-

bration methods have been developed based on pressure rate-of-rise (ROR) pro-

cedures to address the issue. With these methods, though, care must be exercised

to properly account for such influences as temperature, surface conditions, and

gas compressibility and condensation behavior.

The technique of surrogate gas calibration, where a gas with similar molecular

properties is used for the calibration of an instrument for hazardous gas use, has

received some acceptance. However, there is very litde published experimental

evidence that correlates the performance of surrogate gases to actual gases.

386 Chapter

3.4:

Mass Flow Measurement

and

Control

3.4.4

TROUBLESHOOTING

Since an MFC has a set point signal, an output signal, and usually an internal

valve drive signal available at the electrical connector, some troubleshooting is

possible without removing the instrument or otherwise opening the gas supply

line.

In situ trouble shooting can potentially avoid time-consuming and costly re-

placement procedures.

3.4.4.1

Initial Checks

With an appropriate breakout connector and a voltmeter, prepare to measure the

various signals available on the device under test. Measure with respect to the sig-

nal common through a high-input impedance meter.

Verify that the power supply and grounding lines are within the specified

tolerances.

Verify that the output signal is being displayed properly by the readout in-

strument. Check for:

a. Proper settings for range and gas correction factor

Unintentional zero offsets at the display

c. Display voltage calibration

d. Cable losses

3.4.4.2

Operational Diagnostics and Troubleshooting

Ideally, the device under test should have been characterized while working prop-

erly as part of a routine "fingerprinting" exercise for the entire system. A record

of the typical level and variability of each signal becomes extremely useful when

troubleshooting or predicting a potential failure. In addition to measuring the sig-

nals available at the MFC, it is important to have an independent indication that

flow is occurring. This need not be an accurate measurement; only an indication

that there is or is not

flow.

Often a pressure gauge on the vacuum system into which

the flow is directed can serve as the "actual flow" indication.

The troubleshooting chart in Table

and symptom explanation serves as a guide

to possible problems encountered in a gas delivery system. It is worth noting that

some of the potential problems can be remedied without removing components or

otherwise exposing the system to atmospheric conditions.

References

387

Table

MFC Troubleshooting Chart

Set Point

(V)

Actual

Flow

Large

Some

Sensor Output

Large

<5V

OKbul

(V)

oscillating

Valve

Indication

Closed

Open

Closed

Open

Oscillating

Problem

Valve won't close

Control loop problem

Meter problem (false low)

Meter problem (false high)

Valve won't open or no Ap

Valve slighdy clogged or low A;?

Control loop not stable

Valve Won't

Close:

Mechanical

Actuator failure

Out of mechanical adjustment

p too large

Electrical

Valve override applied

Coil failure (N/O)

Connector failure (N/O)

Valve

Won't

Open:

Mechanical

Actuator failure

p too large

Clogged (valve or sensor)

Swelled elastomer plug

Electrical

Valve override applied

Coil failure (N/C)

Connector failure (N/C)

Meter Problem:

False low

Sensor failure (electrical)

Clogged sensor (in bypassed ranges)

False high

Sensor failure (electrical)

Clogged bypass (large ranges)

Thermal syphoning

Control Loop Problem:

• Latch up

• Circuit failure

Control Loop Not Stable:

• Valve adjustment

Orifice size

Plug to seat distance

• Electronic adjustment

Control gain

Control phase

• Gas type

• Line pressure

• Pressure

fluctuations

upstream

• Pressure

fluctuations

downstream

REFERENCES

General and Historical

Benson, J. M., and

C. Baker,

Instrimi.

Control

Syst.,

43 (1970) 85.

Brown, A. E, and H. Kronberger,

Sci.

Instrum.,

24 (1947) 151.

Burggraaf,

Pieter,

Semiconductor

International,

14(3) (1991) 60.

Hawk, C. E., and

C. Baker,

Vac.

Sci.

Technol.,

6 (1969) 225.

Miller, R. W., Flow

Measurement Engineering

Handbook (McGraw-Hill, New York, 1983).

388 Chapter 3.4: Mass Flow Measurement and Control

Olin, J. G., Intecli, 40 (1993) 8.

Sullivan, J. J., S. Schaffer, and R. P. Jacobs, Jr., J.

Vac.

ScL Technol, A7 (1989) 2387.

Tison, S. A., y.

Vac.

Sci. Technoi, A14 (1996) 2582.

Calibration

Clark, R., MKS Applications Engineering Note, 5 (1989).

Le May, Dan, and David

Sheriff,

Solid State Technology, 39(11) (1996) 83.

Mattingly, G. E., Gas flow metrology, NCSL Newsletter, 29(1) (January 1989).

Scott, R. W. W., Developments in Flow Measurement-] (Applied Science Publishers, Englewood, NJ,

1983),

chap. 2.

Widmer, A. E., R. Fehlmann, and K. Ohtani, J. Phys E., 15 (1982) 213.

Sensor Technology

Benson, J. M., W. C. Baker, and E. Easter, Instrum. Control Syst., 43 (1970) 85.

Hinkle, L. D., and C. F. Mariano, J.

Vac.

Sci. Technoi, A9 (1991)

2043.

Komiya, K., F. Higuchi, and K. Ohtani, Rev. Sci. Instrum., 59 (1988) 477.

Segal, R., E. M. Sparrow, and T. M. Hallman, Appl Sci. Res., 1 (1958) 386.

Singh, S.

N.,

Appl. Sci. Res., A7 (1958) 325.

Gas Correction, Time Response, Diagnostics

Dille, J. C, Semiconductor International May 1988) 234.

Gray, D. E., N. M. P. Benjamin, and B. N. Chapman, Microelectronics Manufacturing Technology

(August 1991), p. 35.

Hinkle, L. D., Semiconductor International, 14(3) (1991) 68.

CHAPTER

4.1

Selection Considerations

for

Vacuum

Valves

Neil T. Peacock

MPS Division

ofMKS

Instruments,

Inc.

4.1.1

INTRODUCTION

Most vacuum equipment uses valving for pump isolation, gas flow control, or gas

admission. When selecting a vacuum valve, there are many considerations. Some

of these are whether it is for shutofif or for control, the type of

valve,

the pressure

range over which the valve is to be used, how it is to be actuated, and the seal ma-

terials. Compatibility of construction materials exposed to the process or the abil-

ity of the operating mechanism to tolerate opening against a pressure differential

may also be important.

4.1.2

VALVES FOR SHUTOFF

Valves intended to isolate sections of a vacuum system completely seal off a port

of the valve, and generally are intended to be used only open or closed. In most

^^^•^ All rights of reproduction in any form reserved.

391