Standard

ISA37.16.01

A Guide for the Dynamic

Calibration of Pressure
Transducers

Approved Date

ISA37.16.012002
A Guide for the Dynamic Calibration of Pressure Transducers
ISBN:
Copyright 2002 by ISA The Instrumentation, Systems, and Automation Society. All rights reserved.
Not for resale. Printed in the United States of America. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means (electronic, mechanical,
photocopying, recording, or otherwise), without the prior written permission of the Publisher.
ISA
67 Alexander Drive
P. O. Box 12277
Research Triangle Park, North Carolina 27709

-3-

ISA-37.16.01-2002

Preface
This preface, as well as all footnotes and annexes, is included for information purposes and is not part of
ISA-37.16.01-2002.
This document has been prepared as part of the service of ISAthe Instrumentation, Systems, and
Automation Societytoward a goal of uniformity in the field of instrumentation. To be of real value, this
document should not be static but should be subject to periodic review. Toward this end, the Society
welcomes all comments and criticisms and asks that they be addressed to the Secretary, Standards and
Practices Board; ISA; 67 Alexander Drive; P. O. Box 12277; Research Triangle Park, NC 27709;
Telephone (919) 549-8411; Fax (919) 549-8288; E-mail: standards@isa.org.
The ISA Standards and Practices Department is aware of the growing need for attention to the metric
system of units in general, and the International System of Units (SI) in particular, in the preparation of
instrumentation standards. The Department is further aware of the benefits to USA users of ISA
standards of incorporating suitable references to the SI (and the metric system) in their business and
professional dealings with other countries. Toward this end, this Department will endeavor to introduce
SI-acceptable metric units in all new and revised standards, recommended practices, and technical
reports to the greatest extent possible. Standard for Use of the International System of Units (SI): The
Modern Metric System, published by the American Society for Testing & Materials as IEEE/ASTM SI 1097, and future revisions, will be the reference guide for definitions, symbols, abbreviations, and
conversion factors.
It is the policy of ISA to encourage and welcome the participation of all concerned individuals and
interests in the development of ISA standards, recommended practices, and technical reports.
Participation in the ISA standards-making process by an individual in no way constitutes endorsement by
the employer of that individual, of ISA, or of any of the standards, recommended practices, and technical
reports that ISA develops.
CAUTION ISA ADHERES TO THE POLICY OF THE AMERICAN NATIONAL STANDARDS
INSTITUTE WITH REGARD TO PATENTS. IF ISA IS INFORMED OF AN EXISTING PATENT THAT IS
REQUIRED FOR USE OF THE DOCUMENT, IT WILL REQUIRE THE OWNER OF THE PATENT TO
EITHER GRANT A ROYALTY-FREE LICENSE FOR USE OF THE PATENT BY USERS COMPLYING
WITH THE DOCUMENT OR A LICENSE ON REASONABLE TERMS AND CONDITIONS THAT ARE
FREE FROM UNFAIR DISCRIMINATION.
EVEN IF ISA IS UNAWARE OF ANY PATENT COVERING THIS DOCUMENT, THE USER IS
CAUTIONED THAT IMPLEMENTATION OF THE DOCUMENT MAY REQUIRE USE OF TECHNIQUES,
PROCESSES, OR MATERIALS COVERED BY PATENT RIGHTS. ISA TAKES NO POSITION ON THE
EXISTENCE OR VALIDITY OF ANY PATENT RIGHTS THAT MAY BE INVOLVED IN IMPLEMENTING
THE DOCUMENT. ISA IS NOT RESPONSIBLE FOR IDENTIFYING ALL PATENTS THAT MAY
REQUIRE A LICENSE BEFORE IMPLEMENTATION OF THE DOCUMENT OR FOR INVESTIGATING
THE VALIDITY OR SCOPE OF ANY PATENTS BROUGHT TO ITS ATTENTION. THE USER SHOULD
CAREFULLY INVESTIGATE RELEVANT PATENTS BEFORE USING THE DOCUMENT FOR THE
USERS INTENDED APPLICATION.
HOWEVER, ISA ASKS THAT ANYONE REVIEWING THIS DOCUMENT WHO IS AWARE OF ANY
PATENTS THAT MAY IMPACT IMPLEMENTATION OF THE DOCUMENT NOTIFY THE ISA
STANDARDS AND PRACTICES DEPARTMENT OF THE PATENT AND ITS OWNER.
ADDITIONALLY, THE USE OF THIS DOCUMENT MAY INVOLVE HAZARDOUS MATERIALS,
OPERATIONS OR EQUIPMENT. THE DOCUMENT CANNOT ANTICIPATE ALL POSSIBLE
APPLICATIONS OR ADDRESS ALL POSSIBLE SAFETY ISSUES ASSOCIATED WITH USE IN

ISA-37.16.01-2002

-4-

HAZARDOUS CONDITIONS. THE USER OF THIS DOCUMENT MUST EXERCISE SOUND

PROFESSIONAL JUDGMENT CONCERNING ITS USE AND APPLICABILITY UNDER THE USERS
PARTICULAR CIRCUMSTANCES. THE USER MUST ALSO CONSIDER THE APPLICABILITY OF
ANY GOVERNMENTAL REGULATORY LIMITATIONS AND ESTABLISHED SAFETY AND HEALTH
PRACTICES BEFORE IMPLEMENTING THIS DOCUMENT.
THE USER OF THIS DOCUMENT SHOULD BE AWARE THAT THIS DOCUMENT MAY BE IMPACTED
BY ELECTRONIC SECURITY ISSUES. THE COMMITTEE HAS NOT YET ADDRESSED THE
POTENTIAL ISSUES IN THIS VERSION.
The following people served as members of ISA Subcommittee SP37.16:
NAME

COMPANY

L. Whitby, Chairman
J. Weiss, Managing Director
J. Branom
H. Estrada
M. Montreuil
R. Rhen
R. Staus
T. Vondenbenken
P. Walter
______
* One vote per company.

DeVry Institute of Technology

KEMA Consulting
Branom Instruments Company
BF Goodrich-Advanced Micro Machines
National Research Council Canada
PCB Piezotronics
Pennsylvania State University
Kulite Semiconductor
TCU Endevco Corporation

The following people served as members of ISA SP37 Committee:

NAME

COMPANY

E. Icayan, Chairman
J. Weiss, Managing Director
T. Anderson
C. Flagg
J. Hendrie
A. Mobley
M. Montreuil
H. Norton
R. Staus
P. Walter
L. Whitby
J. Wilson
W. Zubon
______
* One vote per company.

ACES Inc.
KEMA Consulting
SpaceAge Control Inc.
ACS
Lucas Control Systems Products
3M Company
National Research Council Canada
Jet Propulsion Lab
Pennsylvania State University
TCU Endevco Corporation
DeVry Institute of Technology
The Dynamic Consultant LLC
Bently Nevada Corporation

This draft standard was approved for publication by the ISA Standards and Practices Board on
_________________.
NAME

COMPANY

M. Zielinski, Chair
D. Bishop
D. Bouchard
M. Cohen

Emerson Process Management

David N Bishop, Consultant
Paprican
Consultant

-5-

M. Coppler
B. Dumortier
W. Holland
E. Icayan
A. Iverson
R. Jones
V. Maggioli
T. McAvinew
A. McCauley, Jr.
G. McFarland
R. Reimer
J. Rennie
H. Sasajima
I. Verhappen
R. Webb
W. Weidman
J. Weiss
M. Widmeyer
C. Williams
G. Wood

ISA-37.16.01-2002

Ametek, Inc.
Schneider Electric
Southern Company
ACES Inc
Ivy Optiks
Dow Chemical Company
Feltronics Corporation
ForeRunner Corporation
Chagrin Valley Controls, Inc.
Westinghouse Process Control Inc.
Rockwell Automation
Factory Mutual Research Corporation
Yamatake Corporation
Syncrude Canada Ltd.
POWER Engineers
Parsons Energy & Chemicals Group
KEMA Consulting
Stanford Linear Accelerator Center
Eastman Kodak Company
Graeme Wood Consulting

This page intentionally left blank.

-7-

ISA-37.16.01-2002

CONTENTS
Acknowledgments from the Current ISA SP37.16 Subcommittee................................................................ 9
Introduction ................................................................................................................................................. 11
1

Scope................................................................................................................................................... 13

Purpose................................................................................................................................................ 13

Table of Symbols ................................................................................................................................. 13

Transducer properties.......................................................................................................................... 13

4.1

Underdamped second-order systems .......................................................................................... 14

4.2

General transducer properties...................................................................................................... 19

4.3

Properties in the frequency domain.............................................................................................. 20

4.4

Properties in the time domain....................................................................................................... 20

Dynamic pressure generators ............................................................................................................. 22

5.1

Shock tube.................................................................................................................................... 22

5.2

Shockless pressure-step generators............................................................................................ 27

5.3

Pulse generators .......................................................................................................................... 28

5.4

Periodic pressure function generators (sinusoidal pressure generators)..................................... 29

Measurement of transducer properties................................................................................................ 33

6.1

Sensitivity ..................................................................................................................................... 33

6.2

Amplitude response...................................................................................................................... 33

6.3

Phase response............................................................................................................................ 35

6.4

Resonant frequency ..................................................................................................................... 36

6.5

Ringing frequency......................................................................................................................... 36

6.6

Damping ratio ............................................................................................................................... 36

6.7

Rise time....................................................................................................................................... 37

6.8

Overshoot ..................................................................................................................................... 38

Transducer interfaces .......................................................................................................................... 38

ISA-37.16.01-2002

-8-

7.1

Mounting, strain effects ................................................................................................................ 38

7.2

Cavities and passages ................................................................................................................. 39

7.3

Temperature effects ..................................................................................................................... 40

7.4

Acceleration effects ...................................................................................................................... 42

Electronic considerations..................................................................................................................... 43
8.1

Noise............................................................................................................................................. 43

8.2

Cabling ......................................................................................................................................... 44

8.3

Voltage amplifier........................................................................................................................... 45

8.4

Charge amplifier ........................................................................................................................... 46

Data acquisition and analysis .............................................................................................................. 47

9.1

Digital oscilloscope or recorder .................................................................................................... 47

9.2

Data analysis ................................................................................................................................ 48

10

Reporting test results ....................................................................................................................... 48

10.1

Test conditions ............................................................................................................................. 48

10.2

Results and discussion................................................................................................................. 49

11

References....................................................................................................................................... 49

-9-

ISA-37.16.01-2002

Acknowledgments from the Current ISA SP37.16 Subcommittee

This document has a long history. Originally published by the American Society of Mechanical Engineers
as A Guide for the Dynamic Calibration of Pressure Transducers, this document been an ANSI Standard
since 1972. The theory, including the physics and mathematics of dynamic calibration, is timeless, but the
further development of technology has caused changes in some of the methods described in the original
document.
In 1996, the ASME handed the document over to the ISA. ISAs SP37.16 Subcommittee on Pressure
Transducers began editing the document in 1997, in order to update the methods and references while
leaving the timeless aspects alone. The result is ISA37.16.012002, A Guide for the Dynamic
Calibration of Pressure Transducers.
The following SP37.16 Subcommittee members are recognized because of the major involvement they
had with the current revision of this document, and more importantly because of the role they might serve
as reference sources in any future updates: Patrick Walter, Jim Lally, and Bob Goodemote. However, this
does not lessen the contributions of the entire committee in this effort.
As stated in the original abstract: "While not intended as a step-by-step procedure, this document does
contain specific examples and suggested methods for the determination of items of interest in the
calibration of dynamic pressure transducers."
It is quite likely that in a very short period of time, another subcommittee will edit this document to reflect
current technologies and techniques available.
Lawrence Whitby
SP 37.16 Subcommittee Chair

This page intentionally left blank.

- 11 -

ISA-37.16.01-2002

Introduction
The state of art of dynamic pressure calibration and pressure sensor technology has significantly
advanced since the original publication of this document in 1972 (Reference 56). The ASME standard
documents early attempts to develop dynamic pressure calibration methods, some of which never
evolved further into successful technology. Most of the calibration devices described in this document
were uniquely engineered at individual laboratories to meet their specific measurement needs. However,
today, some of these devices have evolved into commercially available products.
The need to measure "nonsteady" dynamic pressure became very important after WW-II during the rapid
development of jet aircraft and aerospace technology. Investigations of turbulence associated with
launch, shock waves upon re-entry, sonic boom, rocket combustion stability, air blast (References 67,
68), and the dynamics involved with weapons testing were significant measurement challenges.
Investigations in these and other areas have necessitated faithful measurement of pressure variations at
6
frequencies from near zero to the neighborhood of 10 Hertz (Hz). The degree of accuracy with which
these measurements must be made varies widely throughout the technical community, as does the use
made of information derived from such measurements. Often there are other complicating factors, such
as severe environmental effects, which must be considered, if meaningful information is to be obtained.
When considering the measurement problem, the investigator must first determine the dynamic
characteristics of the pressure transducer. It is toward the satisfaction of this basic requirement that this
document is directed.
Dynamic pressure calibration at the time this document was originally authored was difficult because of
the limitation of dynamic pressure calibration sources available. Dynamic calibrators were simply not
commercially available. Since then, substantial improvement has been made in the state-of-the-art of
both dynamic pressure calibrators and high-frequency pressure transducers to meet many current
measurement requirements for amplitude, frequency, and accuracy. Most of the dynamic calibrators
available today incorporating fast-acting valves yield dynamic pressure amplitudes that are independently
established. Others use a pressure transducer as a "transfer" standard that the transducer being
calibrated is compared against (References 56, 57).
Although the users requirement for information concerning a transducers response characteristics has
been as varied as the test methods used to obtain the data, current commercial calibrators and digital
data acquisition systems have helped to obtain more accurate information. Unfortunately there have been
many instances where worthwhile data have gone unused because of the manner in which they were
presented. Test reports lacking adequately defined terms, test conditions, or other supporting information
convey little more than misunderstanding to the reader.
The intent of this document is to provide documentation for current techniques and to identify possible
pitfalls associated with the dynamic calibration of pressure transducers. The results of providing such a
document to the technical community will be a better understanding of the basic problems as well as
more effective communication between workers in the field.
This document is not a step-by-step procedure that can be followed without fail to the absolute truth in
pressure measurements. Neither is it an attempt to discuss in detail all of the factors that affect the
accuracy of pressure measurements, e.g., environmental effects, signal transmission, or recording
techniques. References to applicable documents concerning such problems are contained herein, and
Reference 74 deals specifically with the measurement/data acquisition/data utilization process. This
document concentrates on the factors that directly affect dynamic response, such as adapters and
mechanical attachments physically a part of, or relatively inseparable from the transducer, and electronic
equipment that, in practical use, is required for the operation of the transducer. The description of
equipment and techniques appearing in this document will be limited to their use as directly related to
dynamic pressure calibration.

ISA-37.16.01-2002

- 12 -

The clauses of this document are divided into three groups. The first, consisting of Clauses 4, 5, and 6,
discusses the significant transducer properties, dynamic pressure sources available, and the use of
sources to determine the desired transducer properties. The second group, consisting of Clauses 7 and
8, deals with the problems of transducer installation and the immediate electronic signal conditioning
necessary to obtain a satisfactory output signal. The final group, consisting of Clauses 9 and 10, indicates
data-recording methods and recommends procedures for reporting test results.
Although this document focuses primarily on pressure levels above acoustic, it is worth noting that inch
precision condenser microphones, in compliance with ISA 1094-1-4, also have been successfully used for
dynamic pressure measurements on jet aircraft, rocket engines, and other aerospace applications. For
dynamic calibration, the open-circuit sensitivity and the frequency response are normally obtained by a
precision acoustical calibration system (Reference 51) using a pistohphone (Reference 52) for the opencircuit sensitivity, and the electrostatic-actuator method for the frequency response. An acoustical
calibrator (Reference 53) meeting the requirements of IEC 942 (1988) Class 1 may also be used to
calibrate the open-circuit sensitivity.
This document is inconsistent in the use of the word calibration in reference to dynamic testing of
pressure transducers. It should be understood that calibration as used in this document and others
(Reference 1) means a test during which known values of measurand are applied to a transducer, and
corresponding output readings are recorded. The degree of accuracy associated with these dynamic tests
is generally lower, and the manner in which the results are used is generally less rigorous than in the
conventional and more easily controllable field of static pressure calibration.
In preparing this Guide in 1972, the original ANSI B88 Subcommittee on Pressure had considered the
various testing and reporting techniques before recommending specific practices. This present document
represents the first step in the accomplishment of the Subcommittees assignment, which was to develop
or approve standards for the dynamic calibration of pressure transducers in order to improve the quality of
dynamic calibrations.

- 13 -

ISA-37.16.01-2002

Scope

This standard covers dynamic pressure transducers, which are, primarily those used in measurements.

Purpose

This standard establishes guidelines for the preferred techniques and practices in the calibration of
dynamic pressure transducers.

Table of Symbols

a
Ae

gas speed of sound

outlet orifice area

RC
ts

time constant of R-C circuit

settling time

A e
Ai

maximum exit area

t s
tr

shock-wave transit time

Ar
c
d
f

amplification factor
damping
diameter
frequency

temperature
chamber volume
voltage or volume
peak incremental voltage

k
K
l

spring constant
steady-state sensitivity
piston position

T
v
V
V
V

L
m
Ms

inlet orifice area

V
Vp

peak voltage for any cycle

average voltage
peak-output voltage

length of cylindrical passage

Vs

shock-wave velocity

mass
shock wave Mach number

modulation factor
ratio of specific heats of constant

N
number of oscillations
OUT(s)
transfer function
IN(s)
p
pressure
ps
stagnation pressure of supply gas
pa
absolute pressure
P
pressure change
po
equilibrium pressure
p
average chamber pressure

rise time, transducer

pressure and volume

constant

damping ratio
wavelength

rise time, input

frequency in radians per second

ringing frequency

o
r

natural frequency
resonant frequency

Transducer properties

The transducer properties or characteristics of interest to a user will depend to a large extent on the
application involved. This clause defines and discusses some of the properties most often required.
These properties sometimes can be described in terms of the transient response of the device to a step
input, or in terms of its steady-state response to sine-wave excitation, or both.
In defining transducer properties related to dynamic response, the transducer's transfer function provides
valuable information. The transfer function is the ratio of output to input (expressed in the frequency
domain), and forms the basis for the frequency response parameters. Once the transfer function is
known, the input vs. time for any output can be determined. These topics have been treated by those
working in the fields of servomechanisms and network theory, where it is often necessary to describe

ISA-37.16.01-2002

- 14 -

system behavior in both transient and steady-state terms. This approach, using the transfer-function
concept, has much to offer in the consideration of dynamic calibration of pressure transducers.
Because of the limitations of periodic pressure generators, responses to aperiodic pressure generators
must be depended upon to provide much of the needed information on transducers. Measurements of
frequency response using sine-wave pressure inputs are easily defined and understood, whereas the
necessity to convert from the time to the frequency domain makes analysis with aperiodic inputs more
difficult.
Description of pressure transducer dynamic properties is usually based on representation of the
transducer as a linear second-order system with a single degree of freedom, e.g., a simple spring mass
system with damping. Some transducers may be found to be more complex than such a simple system,
and their analysis is correspondingly more difficult. A detailed analysis of a simple single mass and single
spring system follows.
4.1

Underdamped second-order systems

The typical mass-spring mechanical system, which provides the first resonance of a transducer, is
described by a linear second-order differential equation:

(Eq. 4.1)

d2 x
dt

c dx kx f(t)
+
=
m dt m
m

where c indicates the damping, k the spring constant, m the mass, and f(t) the forcing function (provided
by the pressure generator). The use of the Laplace Transform allows the formation of the transfer function
from Equation 4.1.
(Eq. 4.2)

K 2o
OUT(s)
= 2
IN (s)
s + 2 o s + 2o

where OUT(s) is the Laplace Transform of the output,

IN(s) is the Laplace Transform of the input,
K
is the steady-state sensitivity,
o

is the natural frequency of the system in radians per second =

is the complex variable = j (2f ) where f is frequency

is the damping ratio (ratio of actual damping to critical damping)

k
.
m

- 15 -

ISA-37.16.01-2002

The natural frequency is the frequency of free (not forced) oscillations of the sensing element of a
transducer without damping (c = 0). In practical terms it is the measured frequency at which the
o
transducer has a 90 phase shift.
The response of an underdamped second-order system is treated in numerous texts. (See References 2,
3, 4, and 5.) The amplitude and phase response vs. frequency is shown in Figure 1 and Figure 2, and the
time response to a step input is shown in Figure 3.
The model described by Equation 4.2 assumes that the transducer can respond to static pressures
(j = 0) with sensitivity K. Often transducer systems along with their associated electronics do not
respond to static pressures, and in this case, the model can be modified by incorporating the equivalent
of a high-pass RC stage with the transfer function. For example, a piezoelectric transducer with a
dominant mechanical resonance can be approximated by the transfer function
(Eq. 4.3)

K 2o
OUT(s)
s
= 2

2 s + 1 / RC
IN(s)
s + 2 o s + o

Now when s = 0, the response is zero.

Equations 4.2 and 4.3 are written in terms of the variable s and describe the system behavior in the
frequency domain. When the input function is specified, the inverse Laplace Transform can be used to
derive an equation yielding the time domain function. If we apply a pressure step of height A to the
transducer described by Equation 4.2, the time response of the output voltage will be

(Eq. 4.4)

1
e (t) = A 1
e

1 2

d t

1 2

1 2

sin d t + arc tan

where d = o 1 2 .
Equation 4.4 is plotted for various in Figure 3, and yields the time response of the system to a specified
input.
Information on transducer sensors and physical principles can be found in References 2 and 6-12.
Information supporting the preceding analysis can be found in Reference 63.
The above model will next be applied to the description of properties of dynamic pressure transducers.

ISA-37.16.01-2002
10

- 16 -

+20
V = 0.05
V = 0.1

V = 0.15

+10

V = 0.2
V = 0.25

0.3

V = 0.3
V = 0.4

Decibels

Amplitude Factor

V = 0.6
V = 0.8
V = 1.0

V = 0.5

-10

0.1

-20

0.03

-30

0.01

-40

0.5

0.1

0.2

0.3

0.4

0.5 0.6

0.8

1.0

Figure 1 Amplitude response for an ideal second-order system

10

- 17 -

ISA-37.16.01-2002

Phase Shift In Degrees

-20

V = 0.05
V = 0.1
V = 0.15

-40

V = 0.2
V = 0.25
V = 0.3
V
V
V
V
V

-60

-80

= 0.4
= 0.5
= 0.6
= 0.8
= 1.0

-100

-120

-140

-160

-180

0.05

0.1

0.2

0.3

0.4

0.5

0.6

0.8

1.0

10

Figure 2 Phase response for an ideal second-order system

ISA-37.16.01-2002

- 18 1.6

1.4

1.2

10

Amplitude

1.0

V = 0.8
V = 0.6
V = 0.4
V = 0.2

0.8

V = 1.4
V = 1.2
V = 1.0

0.6

0.4

0.2

0
0

4
0

10

t Angle (In Radians)

Figure 3 Response of ideal second-order system to step

input of unit amplitude

- 19 -

4.2

General transducer properties

There are many general properties of transducers that are described in discussions of static pressure
measurements. Those properties, which have specific application for dynamic measurement, are as
follows (see also Reference 1).
4.2.1

Sensitivity

The ratio of the change in the transducer output to a change in the value of the measurand. Generally,
this is expressed in terms of voltage vs. pressure. The sensitivity is represented by the constant K of
Equations 4.2 and 4.3. If the transducer is accurately represented by Equation 4.2, the sensitivity can be
established by static measurements. If Equation 4.3 is the model, dynamic measurement of sensitivity is
required.
4.2.2

Linearity

The closeness of a calibration curve to a specified straight line. Equations 4.2 and 4.3 assume the
transducer output to be linear. Any non-linearity is a deviation from the model equation.
4.2.3

Range

The measurand value, over which a transducer is intended to measure, specified by its upper and lower
limits.
4.2.4

Creep (drift)

A change in output occurring over a specific time period while the measurand and all environmental
conditions are held constant.
4.2.5

Hysteresis

The maximum difference in output, at any measurand value within the specified range, when the value is
approached first with increasing and then with decreasing measurand.
4.2.6

Proof pressure

The pressure that may be applied to the sensing element of a transducer without changing the transducer
performance beyond specified tolerances.
4.2.7

Repeatability

The ability of a transducer to reproduce output readings when the same measurand value is applied to it
consecutively under the same conditions and in the same direction.
4.2.8

Acceleration-Compensation

An accelerometer element internal to the transducer that reduces its sensitivity to motion.
4.2.9

Thermal sensitivity shift

A change in sensitivity of a pressure transducer as a result of a change in steady-state operating

temperature, expressed as % / C or F.

ISA-37.16.01-2002

- 20 -

4.2.10 Resolution
The smallest discernible signal from a measurement system. It may also be referred to as "threshold.
4.2.11 Noise
Any unwanted signal in the measurement system other than the desired pressure response.
4.3

Properties in the frequency domain

Properties in the frequency domain are described by the transfer function of Equations 4.2 and 4.3.
4.3.1

Amplitude response

The amplitude of the transfer function versus frequency (often called the frequency response). It can be
computed from Equations 4.2 or 4.3 by substituting j for s and computing the resultant magnitude.
This plot is frequently normalized (by the sensitivity) to show deviations from a flat amplitude response.
Figure 1 is an illustration of the amplitude response for an ideal second-order system.
The amplitude response contains a great deal of information relative to the transducer, such as resonant
frequencies, bandwidth, and the damping of the resonances.
4.3.2

Phase response

The phase of the transfer function versus frequency. It can be derived from Equations 4.2 and 4.3 by
substituting j for s and computing the phase, as is varied. Figure 2 illustrates the phase response of
the ideal second-order system.
In the time domain, phase influences the instantaneous shape of the response to an input signal and
contributes to a time lag in transducer response.
4.3.3

Resonant frequency r

The measurand frequency at which a transducer responds with the maximum output amplitude. Most
transducers have more than one resonance, and the lowest in frequency, or first resonance, is usually
considered more important. If the first resonance is the dominant one, the second-order system
approximation may be valid. The normalized amplitude response at resonance is governed by the amount
of damping in the system (see Figure 1).
4.4

Properties in the time domain

The properties in the time domain are descriptions of the transducers response to a specified input,
usually a step function.
4.4.1

Ringing frequency (d ) (sometimes referred to as damped natural frequency)

The frequency of free oscillations in the transducer output resulting from a step change in measurand.
The ringing frequency is indicated by the number of oscillations per unit time. For the linear second-order
transducer, the ringing frequency is related to the resonant frequency by

(Eq. 4.5)

d =

1 2
1 2 2

- 21 -

4.4.2

Damping

The energy dissipating characteristic that, together with natural frequency, determines the upper limit of
frequency response and the response-time characteristics of a transducer. In response to a step-change
of measurand, an underdamped (periodic) system oscillates about its final steady value before coming to
rest at that value; an overdamped (aperiodic) system comes to rest without overshoot; and a critically
damped system is at the point of change between the underdamped and the overdamped conditions.
4.4.3

Damping ratio ()

The ratio of the actual damping to the damping required for critical damping. In Equation 4.2, the
coefficient is the damping ratio. Typical dynamic pressure transducers have damping ratios much less
than unity; consequently values of d and r very nearly coincide. The damping ratio is defined
specifically for a linear second-order system. Transducers with more resonances are approximated by
associating a damping ratio with each resonant frequency.
The damping ratio is a useful parameter in both the time and frequency domain. In the time domain, is
related to the amount of overshoot (see Figure 2) and influences the number of ringing cycles present
after a shock excitation. In the amplitude response, is related to the height of the peak at the resonant
frequency.
4.4.4

Rise time

The length of time required for the output of a transducer to rise from 10% of its initial value to 90% of its
final value when excited by a step change in measurand. Rise time is related to transducer frequency
response.
4.4.5

Overshoot

The amount of output measured beyond the final steady output value in response to a step change in the
measurand. The maximum theoretical overshoot of an ideal second-order transducer is 100 percent; this
occurs when is zero. Overshoot is determined as

(Eq. 4.6)

overshoot =

1 2
100e

for the condition 0.1.

4.4.6

Settling time

The time required after the application of a step change in measurand for the transducer output to settle
within a small specified percentage (5 percent) of its final value. For the ideal second-order transducer
with small

(Eq. 4.7)

ts =

3 1 2
d

The settling time increases with smaller and d. The number of oscillations at d required to settle within
5 percent of final value for the ideal second-order transducer is

ISA-37.16.01-2002

(Eq. 4.8)

4.4.7

N=

- 22 -

3 1 2
2

Discharge time constant (DTC)

Time required for a transducer or measuring system to discharge its signal to 37% of the original value
from a step-change measurand. It relates to low-frequency measuring capability for both transient and
sinusoidal events.

Dynamic pressure generators

The dynamic calibration of pressure transducers requires that the measurand produced by a dynamic
pressure generator varies in time in both a known and an appropriate manner. With some generators, the
pressure-time relationship can be predicted quite accurately. With others, the pressure-time relationship
can be established accurately only with the aid of comparison to referenced pressure transducers. While
reproducibility is a highly desirable characteristic of the dynamic pressure generator, it is not an essential
characteristic. When such a characteristic is lacking in a generator, full dependence on the reference
transducer is required.
Dynamic-pressure generators fall into two basic classes: aperiodic and periodic. The aperiodic generators
are characterized by the pulse shapes they produce, such as the step or the peaking pulse. Quickopening valve devices and pulse generators produce pressure rise times generally in the milli-second
range or less. The rise times and the pressure amplitudes generated by these devices vary markedly from
one type of aperiodic pressure generator to another. The shock tube, for example, is capable of
generating pressure steps having rise times in the nanosecond range. A number of the dynamic
calibrators described in this clause are now commercial products.
Pressure step, as used in this document, is defined as a change in measurand in which the rise time is
less than one-fifth the rise time of the transducer measuring it.
Sinusoidal pressure generators, which require the use of a transfer standard, are the most useful of the
various periodic pressure generators available, however, and these devices are limited as to useable
range of frequencydynamic pressure ratio and dynamic amplitude. Nonsinusoidal pressure generators
of significant usefulness include the square wave or rectangular wave generators, which may be
considered as a special case of the aperiodic or step-function generators. Figures 4 and 5 present a
summary of the capabilities of the dynamic pressure generators.
5.1

Shock tube

A shock tube, in its simplest form, consists of two sections of tubing separated by a thin diaphragm. When
these two sections are pressurized to different pressure levels, and the diaphragm is suddenly ruptured,
the higher-pressure gas will immediately begin to flow and compress the gas at a lower pressure
(References 56, 57, 58).
It should be noted that most cold-gas, shock-tube-development work occurred in or before the 1960s.
However, in 1997, a shock tube was designed and built at a university for a transducer manufacturer. The
development report for this new shock tube, Reference 64, also updates the literature through the
intervening time period.
At a distance of approximately 10 to 15 tube diameters downstream from the diaphragm, a well-formed
shock wave is established. This shock wave continues to move through the remainder of the gas in the
low-pressure section at approximately a constant velocity. Behind the shock wave, the pressure suddenly
rises to a new value, resulting in a positive pressure step. The length of time the pressure remains
constant behind the shock wave depends on the dimensions of the shock tube, the position in the low-

- 23 -

pressure section at which the pressure is being monitored, the degree of smoothness of the inner walls of
the low-pressure section, the type and design of the diaphragm, and the type, temperature, and initial
pressure of the gas in each section. Air or helium and air in combination are commonly used gases.
When a shock tube is utilized for pressure transducer calibration, several parameters must be measured
before the amplitude of the pressure step can be ascertained. These parameters include the shock-wave
velocity, Vs, and the initial absolute pressure, P1, and temperature, T1, of the gas in the low-pressure
section.
5.1.1

Sidewall transducer mounting

If a pressure transducer is mounted flush in the sidewall of the low-pressure section, it will sense a
change in pressure, P, when the shock wave passes over it. This is commonly referred to as "incident"
pressure. The equilibrium pressure and particle velocity behind the shock wave are determined from the
Rankine-Hugoniot relations (References 2, 6, 24, 25, 26, 27, 28, and 58). When air is used as the working
gas (low-pressure section), the amplitude of the pressure step can be computed from the following
equations:
(Eq. 5.1)

P =

7
P1 M 2s 1
6

and

(Eq. 5.2)

Vs
Ms =

344.5

298

273 + T
1

where Vs is expressed in meters per second, T1 is gas temperature in degrees, C, P1 is absolute pressure,
and Ms is the shock-wave Mach number. When gases other than air are used, Equations 5.1 and 5.2 do
not apply. (See References 2, 24, and 25 for further information.)
Since the working-gas temperature must be known, the convenient method of determining the
temperature is from a measurement of the static-wall temperature of the shock tube. Except at very low
pressures, the temperature of the working gas closely approaches that of the wall in less than one
minute.
The shock-wave velocity, Vs, is determined by measuring the shock-wave transit time between points
spaced a known distance apart along the path of the velocity vector. Because the velocity of the shock
wave tends to decrease with distance, the last pair of points should be in close proximity (less than 1 tube
diameter) to the transducer undergoing calibration. Several different types of sensors are used to detect
the passage of the shock wave past these points; the most common being pressure transducers, thin-film
heat-transfer gages, and light screens. The shock-wave transit time, ts, between pairs of sensors, is
measured with digital timing. The shock-wave velocity, Vs, is calculated using the equation Vs = spacing
between sensors/t s. Because of the squared relationship between Vs and P, an uncertainty of 0.5
percent in the measurement of the shock-wave velocity produces an uncertainty of at least 1 percent in
the determination of pressure-step amplitude.
When the pressure transducer is mounted flush in the sidewall of the shock tube, the rise time of the
pressure step resulting from the passage of the shock wave is dependent on both the shock-wave
velocity and the transverse length of the transducer diaphragm, d, in the direction of shock-wave
propagation. Consequently, pressure transducers with the fastest rise time and smallest diameter
diaphragm should be used. Commercially available micro-sized pressure transducers with a 1mm
effective area reduce transient time across the diaphragm to 2 to 3 microseconds.

ISA-37.16.01-2002

- 24 -

The time required for the pressure on the transducer to change from P1 to P2 (P2 = P1 + P) is given by the
expression
(Eq. 5.3)

t=

d
Vs

Figure 4 Aperiodic generators

- 25 -

Figure 5 Periodic generators

ISA-37.16.01-2002

- 26 -

The maximum theoretical rise time, tr , for pressure transducers with circular diaphragms mounted flush
in the sidewall of a shock tube can be shown to be
(Eq. 5.4)

t r=

0.687 d
Vs

The side-mounted mode of operation is recommended

1) when this is the manner in which the transducer will be used in application;
2) when maximum accuracy in the determination of the pressure-step amplitude is desired;
3) when it is desirable to minimize transducer ringing; and
4) when the incident wave is considerably cleaner than the reflected wave.
5.1.2

End-wall transducer mounting

If the end of the low-pressure section of the shock tube is sealed off with an end plate, the moving shock
wave, in striking the plate, will reflect from it. This is commonly referred to as a "reflected" shock wave. A
pressure transducer mounted flush in the end plate will detect only the reflected shock wave. The
reflected shock wave is characterized by a much shorter rise time (usually nanoseconds) and a higher
pressure as compared with the incident shock wave (sidewall measurement). The rise time of the
pressure step associated with the reflected shock wave is sufficiently short to excite all the ringing
frequencies associated with virtually all flush-mounted pressure transducers. A tourmaline pressure bar
transducer specially designed for reflected shock-wave measurements has an ~ 0.2 sec rise time and
nonresonant response (References 57, 58). When air is used as the work gas, the amplitude of the
pressure step behind the reflected shock wave is
(Eq. 5.5)

P =

) 25++4MM

7
P1 M 2s 1
3

2
s
2
s

where Ms and P1 are defined as in Equations 5.1 and 5.2.

Because of the complex relationship between P and Ms in Equation 5.5, any uncertainty in the
measurement of the shock-wave propagation Ms may produce an uncertainty in the determination of the
pressure amplitude P several times larger. (Reference 2 provides a convenient source of working
equations when gases other than air are used.)
The pressure behind the incident and reflected shock waves remains constant for a period of time, which
is dependent on the design of the shock tube and on the type, temperature, and pressure of the gases
initially in the two sections. For a given shock-tube geometry, the longest duration of constant pressure
behind the reflected shock wave is obtained when the shock tube is operated under tailored conditions,
as described in References 2, 10, and 26. Depending on the operating conditions and shock-tube
geometry, the period of constant pressure behind the reflected shock wave may vary from a few hundred
microseconds to several milliseconds.
The end-plate mounted mode of operation is recommended
a) for the determination of transducer ringing frequencies;
b) when this is the manner in which the transducer is to be used in application;

- 27 -

c) when the maximum pressure step amplitude is required in calibration;

d) when the maximum duration of constant pressure behind the shock wave is desired; and
e) when determination of ringing frequency of gas passage is associated with transducer recess mount.
5.1.3

Other considerations

Acceleration of the walls and end plate of a shock tube occurs during operation of the device. In order to
determine the effect of this ground shock acceleration on the transducer output, the sensing end of the
transducer must be blanked off from the pressure wave without significantly altering the acceleration
components. Acceleration effects can be minimized by utilizing heavy-walled tubing for the low-pressure
section of the shock tube, by using a heavy end plate, and by shock-mounting the tube. Modern miniature
acceleration-compensated transducers are less susceptible to mechanical vibrations traveling along the
shock-tube walls.
When a shock wave passes through the working gas, the temperature of the gas is suddenly raised. The
new temperature, T2, varies with the square of the shock wave Mach number, Ms, as well as with the type
and initial temperature of the gas. If the transducer is sensitive to transient temperatures, then the
temperature step produced by the shock wave may cause errors in the transducer calibration.
To determine the extent of such errors and to reduce their effects, a temperature shield may be
employed. Thin coatings of opaque insulating materials sprayed on or bonded to the diaphragm make
good shields but may alter the transducer characteristics. A transducer should always be checked for
transient temperature sensitivity (7.3.3 provides further information).
5.1.4

Recommended shock-tube operating conditions

1) When the reflected shock-wave mode of operation is used, the shock tube should be operated under
tailored conditions. (See description in References 2, 26, and 28.)
2) The test transducer should be flush-mounted in the sidewall or solid-end plate of the shock tube
unless special considerations indicate otherwise.
3) The shock tube must be kept free of diaphragm fragments, gas and oil contaminants, moisture, and
thermal effects remaining from previous operations.
4) The significance of both acceleration (ground shock) and transient temperature on the response of
the transducer must be investigated prior to making a pressure calibration.
5.2

Shockless pressure-step generators

A number of devices have been developed that generate a rapidly rising pressure step between two
pressure levels (see References 2, 15, 16, 17, 27, and 58). Most of these units employ a quick-opening
valve. However, at least one utilizes a burst diaphragm. The geometry of the generator and the opening
time of the valve or burst diaphragm are such as to preclude the formation of shock waves when the
device is operated. Shockless step generators have been designed and successfully used to produce
both increasing and decreasing pressure steps. Although most of these generators employ gaseous
media, a few liquid-medium devices have been developed and used. The shockless pressure-step
generator, now commercially available, has the following advantages over other dynamic-pressure
calibrators:
a) The magnitude of the pressure step generated by the device is determined by measurements of static
pressure on the test transducer before and after the quick-opening valve is opened, thereby
permitting high accuracy in its determination.

ISA-37.16.01-2002

- 28 -

b) The duration of constant pressure behind the pressure step can be made as long as desired.
c) Both the initial pressure on the transducer and the magnitude of the pressure step are controllable
over very wide pressure ranges.
d) In general, it is superior to the shock tube from the standpoint of operational speed and simplicity of
technique.
The dynamic characteristics of a shockless pressure-step generator are determined by measurement with
a calibrated reference transducer possessing a rise time of no more than one-fifth that of the measurand.
The following dynamic characteristics of the generator should be known: rise time, overshoot,
undershoot, and the inherent ringing frequencies with their associated damping ratios. Also of interest is
the stability of both static-pressure levels P1 (initial pressure) and P2 (final pressure).
When the pressure rise time is one-fifth that of the transducer undergoing test or calibration, the error in
the measurand value of transducer rise time is less than 1 percent (see 6.8). If this criterion is not met, the
complete rise time must be analyzed carefully for meaningful results.
Acceleration is present during the operation of the shockless pressure-step generator, and this should be
minimized by design. In general, the shorter the rise time of the device, the greater is the level of
acceleration (ground shock). In those units that utilize poppet valves, it may be necessary to open the
poppet valve more slowly when calibrating at very low pressures in order to keep the acceleration level to
a minimum.
Associated with the pressure step produced by these generators is a dynamic temperature change in
which amplitude is related directly to the pressure change, P2 P1, and inversely to the rise time of the
measurand. As with shock tubes, the effect of the dynamic temperature pulse on the response of both the
test and reference transducers must be determined. When a gas medium is used in the shockless step
generator, the rise time of the measurand is inversely related to the speed of sound in the gas. For this
reason helium is used when very short rise times are desired.
The following calibration conditions are recommended:
1) The test transducer should be mounted flush in the generator with a minimum of volume between the
valve or diaphragm and the transducer diaphragm.
2) In the determination of transducer rise time and overshoot, the rise time of the generator should be
less than one-fifth that of the transducer.
3) The amplitude of the pressure steps generated should cover the range of the transducer or range of
application.
4) The medium (gas or liquid) in the step generator should be similar to that used in the final application.
5) When possible, the generated pressure steps should be of the same direction as encountered in the
application, i.e., either increasing or decreasing pressure.
6) The significance of the transducers response to acceleration and to the dynamic temperature pulse
in the generator should be determined.
5.3

Pulse generators

Several devices have been developed to provide single-peaking pulses of reasonably controlled
amplitudes. These pulse generators produce a dynamic measurand that is not a step function, but that
may resemble a single half cycle of a sine wave. One technique employed to generate such a pulse is to

- 29 -

drop a mass onto a piston in contact with the surface of an incompressible fluid contained within a fixed
volume (References 14, 66, and 73). The commercial version of this is referred to as a hydraulic-impulse
calibrator. The device consists of a piston/cylinder manifold and a drop tube containing a mass that can
be dropped onto the piston from various heights. The amplitude of the pulse is dependent on the fluid
incompressibility, the mass, its initial height above the piston, and the piston area. The pulse generator is
not an absolute calibration device and requires a comparison pressure transducer of known
characteristics to monitor the pulse and provide a peak value measurement for the test transducer.
Alternately, commercial versions that operate to 100,000 psi depend on acceleration references on a
known mass (References 55, 64).
The greatest advantage of the pulse generator is the comparative ease with which very high-pressure
pulses can be generated. Care must be taken in the selection and location of the reference transducer
used since results of the calibration are dependent on this comparison standard. Tourmaline transducers,
which are volumetrically sensitive, are commonly used as transfer standards in hydraulic-impulse
calibrators (Reference 57). Hydrostatic pressure is applied directly to the crystal. The recommended
conditions of operation of 5.2 relative to the comparison transducer apply equally well to these
generators. In order to achieve accuracy in calibration using the pulse generator, it is essential that no
pockets of gas exist at the diaphragm of either the comparison or test transducers.
5.4

Periodic pressure function generators (sinusoidal pressure generators)

The dynamic calibration of a pressure transducer could ideally be accomplished by sensing known inputs
from a periodic pressure generator at known frequencies and amplitudes if such a device existed. The
observed response, including the magnitude, waveform, and phase lag could then be compared with the
known input at various conditions. In order to calibrate with only one frequency at a time for accuracy and
simplicity, a sinusoidal pressure generator (SPG) is required. In practice, there are limitations to this
approach. First, the applied average pressure levels and dynamic amplitudes generally are not known by
absolute means, and must be measured by another transducer. The SPG generates a pulsating pressure
in a small chamber that can be monitored simultaneously both by a reference standard transducer and by
the transducer being calibrated (References 21, 57). The two transducers must be sufficiently close so
that they sense the same pressure, including amplitude, shape, and phase lag. Analysis of the output of
the transducer being calibrated is thus entirely dependent on the performance of the reference transducer
and what is known about this performance. The reference transducer, if statically calibrated, should also
be calibrated by dynamic methods to establish that its sensitivity derived from static and dynamic
calibration is the same. Credibility of the dynamic sensitivity of the reference transducer is a basic
limitation of SPG utilization to a comparison process. As long as the reference transducer is provided with
credible dynamic calibration, it may not be a serious limitation because high-quality reference transducers
can be selected that have response characteristics exceeding the pressure, pressure amplitude, and
frequency that can be obtained with available SPGs. Otherwise, it may be difficult to present a compelling
argument concerning the validity of any calibration that uses a statically calibrated transducer as a
reference standard for dynamic calibration.
The governing limitations are associated with the ability of the SPG to provide the desired signal. An SPG
device, when used for calibrating a pressure transducer for a specific use, should satisfy the following:
a) The pressure generated is sinusoidal such that frequencies other than the fundamental are negligible.
b) The frequency range generated covers the frequencies of pressure expected in the intended
application.
c) The operating pressure range covers the transducer rating and/or intended application.
d) The dynamic pressure amplitude generated is large enough to identify possible nonlinearities in the
transducer amplitude response.

ISA-37.16.01-2002

- 30 -

e) The SPG is operated with the same medium (gas or liquid) with which the transducer is to be used.
In many cases, these criteria cannot be met, and a less-than-desired match is obtained between the
dynamic pressure measurand applied during calibration and that encountered in use of the instrument.
Many special devices have been proposed and developed as SPGs, and these are described in
considerable detail (References 2, 18, 19, 20, 21, 22, 23, 67, and 71). The SPGs can be categorized as
acoustic resonators, variable-volume generators, or variable-mass generators. Little has been done to
further develop SPGS, since the 1960s when sinusoidal calibrator research was government funded.
5.4.1

Acoustic resonators

Any of several driving devices can be used to establish acoustic resonance within a chamber. The
frequency of resonance is fixed by the geometry of the chamber and the properties of the working fluid.
To obtain a specific frequency, the chamber length must be a multiple of the acoustic half wavelength.
When another frequency is required, the geometry must be changed, or harmonics of the fundamental
frequency must be used (References 20 and 23).
At resonance, the pressure waves may be distorted to a significant degree from the pure sinusoidal
excitation because of gas dynamic phenomena. As the amplitude or frequency is increased, the
nonlinearities associated with real gas, wall, and friction effects also become significant. At low dynamicpressure amplitudes, these generators can provide pressure pulses that are sufficiently sinusoidal for
many uses.
5.4.2

Variable-volume generators

In the variable-volume generator, a relatively fixed mass of working fluid is alternately compressed and
expanded within a small chamber. The chamber is deliberately made small such that its natural
frequencies are always higher than the frequencies imposed, and thus resonant effects are eliminated. A
piston or diaphragm driver is used to vary the chamber volume and thereby the pressure in a repetitive
manner (Reference 11, 52). This methodology has been mostly used in the development of sound
pressure calibrators.
Usually the gas compression is isentropic, and the pressure, p, follows the piston position, l, as follows:
(Eq. 5.6)

p
l
= o
po l

where po is the equilibrium pressure and lo is the driver (piston or diaphragm) position at the equilibrium
pressure, and = ratio of specific heats of constant pressure and volume (Cp /Cv). Thus, as the piston
travels in a sinusoidal manner, the pressure amplitude is represented by the expression
(Eq. 5.7)

p = po (1 + sin t )

and the dynamic pressure amplitude by the expression

(Eq. 5.8)

p = po p = po [1 (1 + sin t) ] = po sin t

2
p o ( + 1) sin2 t ...
2

where = modulation factor, which is always < 1.

Considering subsequent terms in the expansion with realistic coefficients, this dynamic pressure is clearly
nonsinusoidal. The effects of fluid motion and viscosity also introduce nonlinearities in the wave form.

- 31 -

This class of generators is generally limited by the amplitude-frequency characteristics of the driving unit.
In theory, nonsinusoidal input from the driving device could compensate for nonlinearities and thus a
sinusoidal pressure-pulse shape could be approximated. However, the dynamic characteristics of
practical driver systems (e.g., crystal diaphragm or electromagnetically driven piston) naturally degrade
into sinusoidal displacements as the frequency is increased, thus limiting the effectiveness of this
approach.
5.4.3

Variable-mass generators

The variable-mass generators utilize a fixed chamber volume, and the rate of fluid flow into or out of the
chamber is cyclically varied to develop the dynamic pressure pulsations. These flow-modulated devices
provide a fast response so that relatively larger pressure amplitudes are available at high-frequency
conditions compared to the variable-volume generators. The critical frequency limitation associated with
acoustic dimensions of the chamber still apply, and operation is limited to frequencies appreciably below
the natural frequency of the chamber, which is dependent on the properties of the fluid used and the
chamber dimensions. It should be noted that these devices have not been commercially produced.
A siren-type device is often used for this class of SPG with the gas entering the chamber through a
critical-flow orifice or nozzle from a constant pressure source and leaving the chamber through a larger
critical-flow nozzle (References 22 and 23). The throat of one of these nozzles is interrupted by a rotating
disc or cylinder with throat-sized holes. The pressure in the chamber is dependent on both the rate of
mass addition and the rate of mass discharge.
(Eq. 5.9)

dp
p
ps
= Ai A e
dt
ps

where
+1

(Eq.5.10)

a
=
v

2 2 ( 1)

= constant
+ 1

= gas speed of sound

= chamber volume

= ratio of specific heats of constant pressure and volume

A i = inlet orifice area

A e = outlet orifice area
ps

= stagnation pressure of supply gas.

The dynamic pressure component, P, inside the chamber experienced by both the test and reference
/ 2 (1 + cos t ) :
transducers is given by the following expression when the exit area is A = A
e

(Eq. 5.11)

A
p
1

p = A e
sin t + e cos t + cos 2 t
2
2
4

where A e = maximum exit area and p = average chamber pressure.

ISA-37.16.01-2002

- 32 -

In order that the dynamic pressure component varies sinusoidally, the following conditions must be met:

p=

2A i
A
e

and

A
P
e << 1
p
2
When these conditions are met, the expression for the dynamic pressure component becomes

(Eq. 5.12)
5.4.4

P =

A e p
sin t
2

Fluidic pressure generator

Developments in fluidic devices in the late 1960s and early 1970s were thought to provide the basis for a
new generation of dynamic-pressure generators. Fluidic-pressure generators are typically low-amplitude
devices but are capable of wide dynamic range. Two fluidic-pressure generators are described in
Reference 19, one of which is reported to be capable of low output (0.065 psi) to 105 KHz. Since this
early work, little effort has been applied to develop commercial fluidic pressure generators.
5.4.5

The following are recommended conditions of operation for periodic pressure-function generators:

a) Maximum pressures (average plus dynamic amplitude) should be within the range in which the
reference transducer is known to be essentially linear.
b) The maximum frequency should not exceed one-fifth of the natural frequency of the reference
transducer in order to maintain calibration accuracy to within four percent.
c) The dynamic amplitude should not be so high that appreciable deviations from sinusoidal wave form
should be obtained as indicated by the reference transducer.
d) The calibration should be conducted continuously or with a minimum of ten discrete frequencies
distributed through the range of frequencies of interest.
e) The reference and test transducer separation distance in the chamber should be less than 1/10
wavelength of the highest measurand frequency used (unless phase shift is otherwise taken into
consideration).
f)

The reference and test transducer outputs should be recorded simultaneously.

g) The indicated average pressure and the dynamic pressure amplitude should be determined for each
test point with a reference transducer.
h) Fatigue and thermal effects should be minimized by short-time dynamic operation at each test
condition.
i)

Integral cooling systems normally used in the test transducer should be activated during the
calibrations. Similarly, bleed-type cooling systems should be activated when compatible with the
pressure generator.

- 33 -

j)

The working fluid used in the generator should be the same state (liquid or gas) as in the intended
use of the transducer being calibrated.

k) The reference transducer should be calibrated by other than static methods (i.e., shock-tube and/or
step-pressure generator) to establish dynamic sensitivity and frequency response characteristics.
As noted in 5.4, under periodic pressure-function generators, in many cases the above criteria cannot be
met and a less-than-desired match between the periodic calibration and measurand requirements can be
achieved.

6
6.1

Measurement of transducer properties

Sensitivity

In a transducer sensitivity measurement, either periodic or aperiodic pressure generators may be used to
produce the measurand. It is preferable to use a generator for which the dynamic pressure amplitude can
be accurately established without use of a dynamic reference transducer. To a limited extent, the shock
tube satisfies this requirement. Today, calibration shock tubes with more precise measurement capability
for shock velocity and pressure amplitude are achieving uncertainties approaching 2%. The shockless
pressure-step generators (e.g., quick-opening valve devices) expose the transducer being calibrated to a
precisely known static pressure in about 50 sec (Reference 59). A fixed-displacement piston-phone,
commonly used for calibrating microphones, allows precise sensitivity measurements at low frequencies
without a reference transducer, but only at very low-pressure levels. If the transducer responds to static
pressures, the typical static-pressure generators (such as hydraulic dead-weight testers) can be readily
used to establish the static sensitivity.
When used with a step-pressure source, a transducer with less than critical damping will produce an
oscillatory output. In a shock tube, the reflected wave may disturb the transducer output before the
transducer oscillations decay. In this case, the average value of the oscillations must be estimated in
order to determine the sensitivity. In general, quick-opening valves allow application of an undisturbed
pressure long after the oscillations decay.
If the transducer does not respond to static pressures, waiting for the oscillations to decay can contribute
to an error in measurement of sensitivity. For example, a transducer with a single RC roll-off at 1 Hz
(described by Equation 4.3) has dropped 5 percent of its value approximately 8 milliseconds after the
application of a pressure step. In this case it requires that the oscillations be averaged in the first few
milliseconds for such a transducer. Sinusoidal pressure generators can be used for straightforward
determination of sensitivity; however, precise pressures, as measured by a transfer standard transducer,
can be generated only at low-pressure levels and relatively low frequencies.
Most of the properties defined for transducers indicate the magnitude of sensitivity variation with these
conditions, e.g., variations in response and linearity. It is therefore recommended that measurements to
determine the sensitivity be made under user-operating conditions.
6.2

Amplitude response

Amplitude response is one of the most important (but more difficult to obtain) properties of a transducer.
Ideally, this measurement is performed with a sinusoidal pressure generator, which is swept over the
frequency range, yielding a constant dynamic-pressure amplitude at each frequency. Unfortunately, an
SPG, which covers the amplitude and frequency range for most dynamic transducers, does not exist. In
general, a flat frequency response (constant amplitude) cannot be guaranteed from the sinusodial
pressure generator; therefore, the pressure generated must be monitored by a reference transducer,
which should have sensitivity documented through dynamic calibration techniques. The ratio of the
response of the transducer under test to that of the reference transducer is recorded. The natural
frequency of the reference transducer must be at least five times the measurand frequencies.

ISA-37.16.01-2002

- 34 -

It is difficult to generate sinusoidal pressures at frequencies as high as the first resonance of most
dynamic transducers. Today, dynamic-pressure transducers have frequency response to 500 KHz and
some to >1 MHz. This has led to the use of aperiodic generators (such as the shock tube) to measure the
transducers amplitude response.
6.2.1

Amplitude response measurements by sinusoidal pressure

Limitations of sine pressure generator measurements are discussed in this subclause. The reference
transducer must be located very close to the transducer under test so that the same pressure field is seen
by both transducers. A general rule is that the distance separating the transducers be less than a tenth
wavelength of the pressure wave. The pressure wavelength is
(Eq. 6.1)

a
f

For air, the speed of sound, a is 1087.4 ft/sec at 0C and at one standard atmosphere, which results in a
wavelength of approximately 1.3 inches at 10,000 Hz. The value a increases as the square root of the
absolute temperature and is essentially independent of pressure.
The geometry of the setup is critical for high-frequency measurements. Typical sinusoidal pressure
generators operate into a sealed cavity. The dimensions of this cavity must be such that it does not have
resonances in the frequency range of interest. For reasonable data, the first half-wave cavity resonance
should occur at least five times higher than the highest test frequency. For air at room temperature 0.1inch cavity length would resonate at approximately 33,000 Hz, indicating a usable frequency range to
6,600 Hz. It must also be established that the pressure wave is essentially a plane wave; that is, that
resonant modes do not exist across the cavity.
The usable frequency of a cavity is often increased by changing the media from gas to liquid. The sound
propagation velocity C in many liquids is approximately five times that in air. This increases the usable
frequency of the 0.1-inch cavity to 33,000 Hz, assuming a cavity with rigid walls and fluids with no gas
bubbles.
If the ambient environment changes from a gas to a liquid, a question is raised about the validity of
calibration with a liquid of a transducer to be used in a gas. Equation 1.2 shows the transducers transfer
function to be dependent on the transducer parameters of stiffness, mass, and losses. It is apparent that
the equivalent moving mass would be greater for a liquid than a gas, and the losses could be different.
Changes in these coefficients may alter the resonant frequency and the damping ratio considerably.
Harmonic distortion in the sinusoidal pressure generator can indicate false minor resonant frequencies in
the amplitude response. For example, if a resonance exists at f1 (either from transducer or cavity), and the
th
generator being used has a strong n harmonic distortion, a minor resonance may falsely appear when
th
the generator is at f1 / n. For this to occur, the resonant amplification times the n harmonic must be equal
to or greater than the fundamental; for example, 20 dB resonance and a 10 percent second harmonic. If
this effect is suspected in a measurement, the minor resonances can be examined by rejecting the
harmonic in the transducer output by the use of a tunable filter. In this way it can be demonstrated that
minor resonances are valid and not due to a testing error.
The influence of distortion of the pressure waveform on the accuracy of the amplitude response depends
upon the method of reading the output signals. For example, if the peak-to-peak value of the
instantaneous signal is taken from a digital recorder, distortion and noise can disturb the shape of the
wave and make the readings difficult. However, if root-mean square or rectified-average readings are
used, small perturbations on the instantaneous signal are not significant. For this reason rms or rectified
detection is recommended for all steady-state sinusoidal measurements.

- 35 -

It is sometimes necessary to acquire response data at pressure levels that are low relative to the
transducers full-scale capability. In this case, electrical noise may be a problem, and it may be necessary
to use a filter. The characteristics of this filter may be determined in the following manner by inserting a
voltage at the transducer at the same frequency. A low resistance is inserted in the ground return lead of
the transducer, and an accurate AC voltage is applied across the resistor from an ungrounded oscillator.
The insert voltage-vs.-frequency characteristic can be established with high accuracy, and this is used to
correct the pressure data to yield the curve for the transducer alone. As an alternative, the entire system
(transducer and filter) can be calibrated as a unit using the techniques previously outlined.
6.2.2

Amplitude response with aperiodic sources

The shock tube can be used to determine the amplitude (and phase) response of a transducer by
mathematically transforming the pressure input and transducer output from the time domain into the
frequency domain. The theory for the mathematical operations is fully developed and can be quickly
performed with digital signal-processing instrumentation (Reference 64).
It is typically assumed that the pressure input to the transducer approximates a step function. A reference
transducer is not used. The response to the step function depends upon whether the test transducer is
subjected to reflected measurements at the end or along the side of the shock tube. The rise time of the
reflected pressure at an end-mounted transducer is very fast, but a side-mounted transducer sees the
incident shock wave pass across its diaphragm, resulting in a much slower rise time. In this case the
sensitivity of portions of the diaphragm become significant, and the fine structure of the input is difficult to
determine. The frequencies determined by this technique may be limited by the pressure rise time of the
shockwave and the duration of the pressure plateau.
In dealing with the transient response, care must be taken that other inputs (from temperature or
vibration) do not contribute significantly to the transducer output. Otherwise transducer-mounting
resonances may appear in the pressure data (see Clause 5). The geometry of inlet coupling ports
associated with recess-mounted transducers also affects their amplitude and phase response.
The time between samples limits the upper frequency that can be resolved. Typically, experimenters
prefer 5 or more samples per sine wave for wave-shape definition. For example, a data sample every five
microseconds would allow wave-shape approximation to 40,000 Hz. However, frequency information
would be preserved to 100,000 Hz based on Nyquist sampling. Given a transducer response that is clean
for five milliseconds, a total of 1000 data points would be generated.
6.3

Phase response

Phase response is determined with a sinusoidal pressure generator by comparing the transducer output
voltage waveform with that of the reference sensor, which is simultaneously excited by the periodic
measurand. The apparatus is much the same as for amplitude response measurements, except the
phase difference between the two waveforms is determined by means of a phase meter, Lissajous
patterns, or by accurately measuring the time shift between the two wave forms of a digital recorder; if not
dual beam, operate in chop, not alternate, mode.
The reference-pressure transducer should have negligible phase shift in the frequency range in which the
test transducer is to be measured. Figure 2 shows the phase shift for second-order systems with various
degrees of damping.
The positioning of the transducers in the measurement cavity is more critical for phase measurements
than for amplitude measurements. A criterion of placement of /10 was suggested for amplitude
measurements, but this is equivalent to a 36-degree phase shift. Hence the criteria for phase-shift
measurements should be increased to /50 to insure phase differences of no greater than 5 degrees. This
is a very stringent requirement and essentially limits phase measurements to low frequencies.

ISA-37.16.01-2002

- 36 -

A technique for circumventing these criteria is to place the reference and test transducers at mirror
image locations within the cavity. The use of two reference transducers is required to establish to what
extent the locations are mirror images.
In general, electronic filters are not recommended for use in phase measurements unless the cut-off
frequencies are a decade above the measurement frequencies, or unless the filters used for the
reference and test transducers are matched in phase. The use of the insert voltage techniques, to
determine the validity of the method, is recommended if filters are used (Reference 10).
Phase measurements can also be mathematically derived from shock-tube measurements as a byproduct of the amplitude-response data.
6.4

Resonant frequency

The transducer-resonant frequency is best determined from the transducers amplitude response,
obtained with a reflected shock-tube measurement. Sinusoidal generators do not have high enough
frequency to excite the resonant frequency of most dynamic-pressure sensors. It is readily determined by
examination as the frequency at which the transducer responds with maximum-output amplitude.
Resonant frequency can be measured quickly and accurately using computer-data-processing software.
The resonant frequency (r) can also be calculated from the ringing frequency (d) by use of the
relationship

(Eq. 6.2)

r =

1 2 2
1 2

Because the resonant frequency is a function of damping ratio as well as natural frequency, the type of
pressure medium is important since the damping effect of a liquid medium can significantly affect the
resonant frequency.
6.5

Ringing frequency

The response of an underdamped transducer to a step or impulse is a damped oscillatory transient or

ringing. The ringing frequency can be determined by applying a pressure-step input, usually from a
reflected shock wave to the transducer, recording the output and counting the cycles-per-unit time of the
transducers response. Typically, these data are quickly recorded and analyzed on a high-frequency
digital recorder. When a transducer exhibits more than one ringing frequency, the output is a combination
of these frequencies, and the measurement is determined with computer software. In this case the
various ringing frequencies must be separated by means of filters before Equation 6.3 can be applied. It
can be shown that the frequency response of such a system, below the lowest ringing frequency, is quite
similar to that of a single-degree-of-freedom system.
6.6

Damping ratio

The damping ratio () can be obtained from the amplitude response using a sine-wave generator or can
be measured using an aperiodic generator.
The amplification factor (A r) of a resonance on the amplitude-ratio curve is related to the damping factor
(assuming a linear second-order system)
(Eq. 6.3)

Ar =

1
2 1 2

; 0 0.707 = 1 ; 0.707 1

and is plotted in Figure 1. Solving for :

- 37 -

(Eq. 6.4)

2 =

1 1 1 / A r2
2

If a step input is used to excite the transducer, the damping ratio can be determined from the following
relationship (assuming a second-order system):

(Eq. 6.5)

2





2N

= 1 +

V1

2.303 log10 V
2

1/ 2

where N represents the number of complete cycles at a specific ringing frequency over which the
measurement is made; V1 and V2 are the peak incremental voltages above the average value at the
beginning and end of N cycles.

The accuracy of this method depends on the ability to establish the average value of voltage V , since
V = V V where V is the peak voltage for any cycle. Because of the uncertainty in V , the greatest
accuracy in the determination of is achieved when N equals 1. Equation (3.6) then becomes

(Eq. 6.6)

6.7

2




2.728

= 1 +

V1

log10 V
2

1/ 2

Rise time

Rise time is measured by applying a step input to the transducer and measuring the time required for
output to go from 10 percent to 90 percent of the final average value. For values of damping ratio of 0.5 or
less, the rise time of the step-pressure input must be less than one-fifth that of the transducer for the
transducers rise time to be within 1 percent of its asymptotic value. So long as the rise time of the step
pressure input is less than one-fourth that of the transducer having damping ratio of 0.1 or less, the rise
time of the transducer will be within 1 percent of that obtained with a step-function (zero rise time) input.
Care must be exercised that the rise time of the recording system is sufficiently short to introduce
negligible error in the measurement.

ISA-37.16.01-2002

6.8

- 38 -

Overshoot

Overshoot is measured by observing the transducers response to a step input of pressure. The
overshoot is determined as the peak output (Vp) minus the average output ( V ) divided by the average
output

(Eq. 6.7)

Vp V
100 in percent
Overshoot =
V

The maximum theoretical overshoot that a linear second-order system can have is 100 percent. This
occurs when the damping ratio is zero. The maximum overshoot is strongly dependent on both the
damping ratio and the quantity d where is the input rise time, and d is the ringing frequency of the
transducer. Most pressure sensors have damping ratios less than 0.1. Acceleration-compensated
pressure transducers have substantially less overshoot than non-compensated transducers. In addition,
their overshoot is not linear with step-pressure amplitude. Generally, but not in all cases, their percentage
overshoot will increase with larger step-pressure amplitude.

Transducer interfaces

The following four main factors need to be considered when installing dynamic-pressure transducers
either for calibration or for performing a measurement:
a) Strain effects
b) Cavity or passage resonances
c) Temperature effects
d) Acceleration effects
Other effects, such as earth-gravity field and those from the earth's magnetic field, which may be
significant when dealing with larger, more delicate mechanical instruments, will usually not affect the
dynamic behavior of an electrical-pressure transducer but should be checked for in some cases. When
calibrating vibration transducers on a shake table, spurious effects induced by the strong variable
magnetic field of the table are sometimes encountered. Again, such effects are seldom significant in
dynamic calibration of pressure transducers and will not receive consideration in this clause.
7.1

Mounting, strain effects

Both non-precision mounting and over-torquing induce strain into a transducer housing (Reference 73)
and can be a source of measurement error. Strain introduced into the transducer body may manifest itself
either as a change of the sensitivity, an increase in mechanically induced noise, or as a null shift. Strain
sensitivity shift is normally noticed when calibration data varies in slope, depending on the mounting
torque applied. In order to assess this effect, the transducer should be calibrated first using the
recommended mounting torque and then repeated with some specific over-torque and under-torque
values, respectively. Calibrating a torque-sensitive transducer satisfactorily is a difficult task since other
factors such as concentricity of the mounting hole, tightness of the thread, dirt particles, etc., may affect
the sensitivity and thus yield inconsistent results, which are hard to correlate.
The null shift (zero shift) caused by mounting strain represents the component of the transducer signal,
which does not depend on the input pressure, but which is a shift in location of the calibration data curve.

- 39 -

A torque-sensitive transducer acts like a strain gage. In a dynamic calibration or measurement, it will
generate a spurious signal that reflects strain fluctuations in the mounting structure. No standard
technique has been defined yet to accurately determine the magnitude of this effect in pressure
transducers. A simple method of detecting significant strain sensitivity is to connect the transducer to its
recording equipment and check the change of the output while tightening it to its recommended mounting
torque in its mounting hole. This technique is useful only if the system has a good low-frequency
response. Manufacturers installation drawings should be followed closely and only the mounting parts
should be machined according to dimensions provided.
Transducers should be installed using the recommended mounting torque. Care must be taken in order
to avoid strains in the mounting structure that can affect the transducer performance during calibration
and use.
7.2

Cavities and passages

The manner in which a transducer is mechanically coupled to the pressure can significantly affect the
response of that transducer. Meaningful measurements of pressure fluctuations at frequencies around
10,000 Hz or higher can only be made with transducers having flush diaphragms. The use of any
connecting line or cavity will limit the frequency characteristics of the measurement system itself.
There are instances where a connecting line or passage cannot be avoided. In such a case, its length will
have to be selected to be consistent with the highest frequency to be measured. The lowest longitudinal
resonant frequency of a cylindrical passage is
(Eq. 7.1)

f=

a
4L

where a is the speed of sound in the gas at the given temperature and L is the length of the passage.
If a dynamic measurement or calibration has to be performed through a passage, the highest frequency
considered should be less than 1/10 of the resonant frequency of the passage. In air at room temperature
(a = 1087.5 feet per second) and a passage length of 1/4 inch, for instance, frequencies up to 1400 Hz
will result in less than one percent dynamic error. This relationship applies only for straight passages. A
small passage leading to a cavity in front of the transducer will result in much lower-resonant frequencies.
The following relation shows the effect of a gas-filled passage and a cavity
(Eq. 7.2)

f=

cd
[(L + 0.85d) V]1/ 2 providing d 2L < 4V

7.09

where
d = inside diameter of tube (orifice) inches,
L = length of tube (orifice) inches,
V = cavity volume, cubic inches.

4V
the cavity resonance approaches the longitudinal resonant frequency described by

Equation (7.1).
When d 2 L >>

The determination of all mechanical resonances basic to the transducer over the frequency range to be
measured is important. The nature of these resonances may be somewhat obscure, and care must be
exercised to insure their repetition from one installation to the next if a response analysis is to be valid.

ISA-37.16.01-2002

- 40 -

Resonances other than the major ringing frequency may sometimes be caused by non-flush diaphragms,
discontinuities in the surface near the transducer, and vibration, etc. In short, modulating frequencies
above or below the ringing frequency may not be inherent to the transducer at all. The presence of these
associated resonances emphasizes the fact that if evaluations of transducer response are to be
meaningful, the mounting configuration employed for the calibration must be identical to that used in the
actual application. Liquid-filled systems may introduce additional measurement uncertainties due to
cavitation and inertial effects that are not readily predictable.
7.3

Temperature effects

Because of the temperature sensitivity of many dynamic-pressure transducers, temperature effects need
to be considered in most applications and in calibration as well. When calibrating a temperature-sensitive
transducer with compressed air, for instance, the small temperature rise due to adiabatic compression
may be sufficient to significantly distort the results. There are two basic effects due to temperature, the
temperature-sensitivity shift and the temperature-null shift. Furthermore, unevenly distributed (i.e.,
transient) temperatures may cause quite different effects from those obtained with the transducer heated
uniformly in a lab oven (Reference 31).
Accurate calibration of a dynamic transducer can only be obtained after its temperature sensitivity has
been assessed. If that should turn out to be excessive, temperature should be maintained constant at a
predetermined value during calibration.
There are at least two methods for evaluating the transient thermal sensitivity of pressure transducers.
The easiest method is to make a test measurement with and without ablative coating applied to the
transducer diaphragm (see 7.3.3). This may be practical if the cost factors involved with the test
measurement are not significant and the data is repeatable, as might be with periodic compressor
measurements. If the periodic measurand is not repeatable, and if practical, two transducers can be
usedone with and one without ablative diaphragm coating. One advantage of this evaluation method is
that both transducers are exposed to the actual heat flux associated with the measurand. No differences
in the two measurements would indicate thermal transients are not a problem.
Another method used for evaluating the thermal sensitivity consist of intermittently exposing the
transducer diaphragm to a heat source. This test can be accomplished by mounting the transducer on a
plate and only exposing the diaphragm to a heat source, which is quickly passed across the transducers
diaphragm, while observing or recording the output. For comparative evaluation, the test can be
performed with and without an ablative diaphragm coating. A variable-speed rotating disk with a slot or
square hole in it can be used to better control the heat exposure time and repeatability, This disk is
rotated between a heat source and the transducer diaphragm. When the slot or hole in the rotating disk
passes in front of the transducer's diaphragm, the diaphragm is exposed to the heat source. A torch or
heat lamp can be used to evaluate the effects of various wavelengths from convective or radiant heat
sources. If the rotating disk is constructed with multiple slots and driven by a variable speed drive, it may
be possible to adjust heat exposure to the transducer's diaphragm in order to approximate the heat
exposure of the measurement application. In order, to quantify and control heat flux, a heat-flux
transducer can be installed in place of the pressure transducer, and the distance between the heat source
and the heat-flux transducer can be adjusted accordingly.
Dynamic transducers having large temperature-null shifts can be more precisely calibrated using a liquid
medium (water or oil). Care must be taken in order to avoid the creation of an air bubble in front of the
diaphragm, since the air bubble will generate some heat when compressed. When calibrations are
performed with a gaseous medium, the temperature null shift must be checked by subjecting the
diaphragm to a sudden heat input and measuring the signal thus generated.
Three special transducer configurations are used to protect dynamic-pressure transducers from extreme
temperatures as encountered in rocket engines, and in internal combustion engines. These are

- 41 -

a) liquid-cooled transducers
b) gas-bleed transducers
c) transducers with diaphragm coating
These configurations may affect calibration results and may require special calibration techniques.
7.3.1

Liquid-cooled transducers

There are two different approaches to the problem of cooling a pressure transducer: intrinsic cooling and
use of a cooled adapter.
The intrinsically cooled instrument features cooling passages inside the transducer body, sometimes
even crossing through the diaphragm. In other designs, the cavity behind the diaphragm is filled with
circulating, cooling fluid.
The pressure of the cooling fluid may affect performance of the transducer in two ways: by changing its
sensitivity and by shifting its null. The latter effect is usually by far the predominant one. When calibrating
an intrinsically cooled transducer, both effects should be checked by performing calibrations with different
coolant pressures.
Dynamic characteristics may be substantially affected through the presence of the cooling fluid. Liquidfilled cavities will tend to resonate at low frequencies and with large amplitudes. Dynamic calibration
should therefore be conducted both with and without transducer coolant flow.
Some modern miniature transducers are too small to provide space for liquid tubings and fittings, etc.
Those transducers are normally cooled by mounting them into a larger liquid-cooled adapter. For most
efficient cooling, the adapter is often designed to almost completely surround the miniature transducer,
leaving only a small passage in front to couple the pressure to the diaphragm. These indirectly cooled
transducers are practically unaffected by coolant pressure. However, the coupling passage may present
problems in dynamic calibration as discussed in the previous paragraphs.
It should be noted that liquid cooling, whether intrinsic or indirect, does not eliminate temperature effects
in applications involving high rates of heat transfer. Many transducers are very sensitive to the high-heat
flux resulting in such applications, and, in spite of liquid cooling, pressure sensitivity can shift drastically
(see References 32 and 33). There is no presently accepted method for calibrating transducers at
measurand pressures under heat-flux conditions. Dynamic measurements made under those conditions
may be subject to considerable error. Subclause 7.3 provides more information on transient thermal
testing and 7.3.3 for diaphragm ablative coatings.
7.3.2

Gas-bleed method

A design developed for measuring dynamic pressures under the most extreme temperature and heat
transfer conditions as experienced in rocket engines makes use of a passage filled with gas of low
molecular weight (helium). A constant flow of cool gas is bled into the cavity in front of the diaphragm
from a high-pressure supply. (See References 22, 24 and 29).
This arrangement has several advantages. The cool-bleed gas protects the diaphragm from the hotcombustion gases, keeping the transducer element in a moderate environment. The problem of passage
resonances is minimized because of the high-resonant frequency and better damping characteristics of
the low-density gas.
Gas-bleed transducers do, however, have some limitations: they have a limited useful pressure range
between 1/20 and 1/2 of the supply gas pressure if critical-flow orifices are used to control the coolant

ISA-37.16.01-2002

- 42 -

flow rate. If the dynamic pressure to be measured is too low, the speed of the gas escaping through the
passage becomes too high, and a substantial pressure drop develops between diaphragm and
measurand pressure. Therefore the relationship between the pressure upstream and downstream of the
passage is no longer linear.
Gas-bleed transducers are calibrated with gas pressure only, since water or oil would contaminate and
block the extremely small-bleed passages. Since a constant flow of bleed gas would disturb the setting
for precision calibrating conditions, static calibration is usually performed with the bleed-supply duct
capped. Dynamic calibration is usually limited to a shock-tube test with or without bleed-gas flow. The
shock tube has proven to be a practical tool to determine the best bleed-flow rates.
7.3.3

Diaphragm coatings

For relatively short time measurements, a thin coating of silicone rubber or similar material over the
diaphragm is often used (References 32, 33, and 34). Thickness of the coating should be about ten to
twenty thousandths of an inch (for 1/4-inch diameter diaphragm). The coating should be opaque in order
to shield off heat transfer by radiation. GE number 106 Red/Rouge Silicone Rubber & Adhesive Sealant is
a commonly used, durable, high-temperature protectant. Thin, reflective, aluminized mylar tape is
effective for use in radiant-heat environments associated with shock waves. A single layer of common
black plastic electrical tape is effective for convective flame associated with a periodic explosivecombustion measurement, although it may need to be replaced frequently. Some high-pressure ballistic
transducers incorporate a captive pocket in the diaphragm for thermal protectants; others have a semipermanent flame-sprayed ceramic coating.
Applied properly on gages with low-compliance diaphragms (piezoelectric or similar types), the coating
has little effect on gage performance except at very low pressures (e.g., less than 10 psi). Resonance
frequency may be only slightly lowered, and sensitivity and hysteresis are not noticeably affected.
Therefore, the calibration data obtained from the uncoated gage may be directly used for evaluating test
results. However, it is highly desirable to calibrate the transducer with the coating applied, especially for
low-level measurements in the range of small psi values. In order to avoid undesirable contamination or
softening of the coating, gases should be used preferably as calibrating media. If the use of a liquid is
desirable, tests should be made to prove that the coating will not be affected by the liquid.
7.4

Acceleration effects

Dynamic-pressure phenomena are often accompanied by acceleration effects. In a sinusoidal pressure

generator, for instance, the pressure variation may cause vibrations of the chamber wall, which in turn
may induce acceleration and strain effects in the transducer housing. Transducers are often sensitive to
these effects and may generate spurious signals, which are superimposed upon the actual pressure
signal. Such errors can sometimes be quite substantial, comprising a large portion of the transducer
output signal, especially at low-pressure levels in high-vibration environments. With an accelerationcompensated transducer, these effects will be much less severe but may not be completely absent. A
transducer that is acceleration-compensated along its most sensitive axis will generally display some
acceleration sensitivity along some cross axis. Strain sensitivity may add some additional efforts as
mentioned in 7.1.
In order to assess the magnitude of the acceleration error, the following procedure is normally used: The
transducer is mounted in a blind hole in a similar location and similar fashion as the original mounting.
Caution is used caution so that the transducers diaphragm does not bottom out in the blind hole. Thus, it
is presumably exposed to the same vibration (and possible strain) effects, without being subjected to the
pressure itself. The size of the error signal can be directly measured. Care needs to be taken to insure
proper venting of the volume in front of the transducer. Because of a possible phase shift between the
pressure signal and the acceleration signal, determination of the true pressure signal through subtraction
of the two signals obtained may not be readily accomplished. When operating transducers in high-

- 43 -

vibration environments, acceleration-compensated designs will minimize the sensitivity to vibration by an

order of magnitude compared to non-compensated designs.
In a dynamic calibration with a shock tube, the vibration effects may be especially severe. The large and
very abrupt pressure change taking place in the driver end causes compression and expansion waves to
travel along the metal walls. These waves may subject the wall and any instrument mounted in this wall to
very large accelerations. Modern, miniature, low-mass, acceleration-compensated transducers are much
more immune to these effects (Reference 63).

Electronic considerations

Once the measurand has been converted from a physical stimulus to an electrical signal by the
transducer, the signal must be conditioned for transmission to some data acquisition device. Recovery of
the signal depends on the compatibility of several subsystems: namely the transducer, the cabling, the
signal conditioning, the power supply, and the data acquisition equipment. References 35-42 provide
detailed information.
The signal conditioning equipment associated with calibrations employing aperiodic measurand should
have a rise time less than 1/5 that of the transducer being calibrated if it is to faithfully transmit the
transducer output. The frequency response of the signal conditioning electronics associated with
calibration using periodic signals should be such as to introduce negligible error at the highest frequency
of calibration.
8.1

Noise

System noise can become a problem when the amplitude of the output signal of the transducer is low.
Noise can originate in the transducer, the cabling, the signal conditioning electronics, the power supply,
and the recording equipment. With any noisy system, the calibration signal can become lost in the noise
(see Reference 35).
8.1.1

Transducer noise

Pressure transducers can be susceptible to spurious mechanically excited accelerations, temperature

changes, and electromagnetic fields. Clause 7 provides a further discussion of spurious transducer
signals due to improper interfaces. Except at temperature extremes, acceleration-compensated pressure
transducers with integral electronics which provide a high-voltage, low-impedance output are more
immune to these effects than charge-mode transducers with a high-impedance output.
8.1.2

Cabling noise

Cables can contribute to the noise in a measuring channel either by picking up external magnetic and
electrical fields or by generating noise signals internally. External noise may be introduced into the
system through imperfect shielding or through ground loops. Internal noise may be generated through the
tribo-electric effect when the cable is being moved or stressed (References 38, 60, and 71). This consists
of charges generated through friction between the insulator and the shield or between the insulator and
the inner conductor. The higher the impedance of the circuit, the more it is susceptible to this type of
noise. Piezoelectric charge-mode systems are, therefore, especially susceptible to disturbances from the
tribo-electric effect. Transducers that have a very low-signal output, like strain gages, are most sensitive
to disturbances from external magnetic and electric fields and ground loops.
Integral Electronic Piezo Electric (IEPE) transducers are relatively immune to cable effects previously
listed above (Reference 54). A significant advantage of the IEPE transducer is its capability to operate
with long standard coaxial cable, or ribbon wire, without the loss of signal strength or an increase in noise.
This not only improves the transducer and system performance characteristics but also greatly reduces

ISA-37.16.01-2002

- 44 -

the cost of cable. IEPE transducers operate from a constant-current power source. Low-noise cable and
charge amplifiers are not required.
8.1.3

Amplifier noise

All amplifiers have a noise figure associated with them. This figure is usually the noise as referred to the
input. This noise figure is constant and independent of any input cabling.
8.1.4

System noise

System noise includes those effects previously listed as well as noise introduced into the system from
external sources by inductive and capacitive coupling. Inductive coupled noise exists in a system when
multiple ground loops are present. In this case, ground-loop current can flow, due to different ground
potentials, in the signal leads of the system. By using proper grounding and by appropriate use of
differential amplifiers, this type of noise can be reduced. One-point grounding is recommended, preferably
at the recording station. Even in a properly grounded system, the effects of stray capacitance cannot be
entirely eliminated, but they can be minimized by using high-quality shielded cable and by grounding the
system at only one point. References 75 and 76 provide more information.
8.2

Cabling

In practice, the transducer output is transmitted to the signal-conditioning equipment through some form
of cabling. With bridge-type or IEPE transducers, the cabling is also used to supply electric power to the
transducers.
When calibrating a transducer with certain older-type voltage amplifiers, or newer separate miniature
solid-state, source-follower amplifiers, it is desirable to us cable having similar length and characteristics
to that intended for actual usage; and/or a cable of a type recommended by the transducer manufacturer.
This is not a requirement with IEPE transducers, which do not require special cabling, nor do they require
calibration with the same type and length cable intended for the application.
Cabling used with charge output piezoelectric transducers should be high-quality, low-noise shielded type
(Reference 60). Cable length becomes important if older voltage or source follower amplifiers are used,
since long cables increase the capacitive loading, thereby decreasing the signal. In these systems
(Eq. 8.1)

Vo = Q C

where Vo = ouput
Q = charge
and

C = capacitance

Charge amplifiers are more tolerant of cable length but do experience an increase in noise output as a
function of cable length. Low-noise cables must be kept clean and dry in order to prevent lower resistance
from causing drift and the loss of low-frequency response. This becomes a limitation when the application
requires operation in high humidity or other contaminated environments.
When measurement applications require operation through long cables in adverse environments, IEPE
transducers offer distinct advantages. They will drive long cables without loss of signal strength or an
increase in noise. However, the capacitive effect of very long cables may act as a low-pass filter. Some
IEPE pressure transducers operate with a high current (20 ma) as to provide cable-driving capability.
Cable-driving monographs can be found in some manufacturers catalogs.

- 45 -

When the Guide was originally authored, some transducers required given lengths of cable between the
transducer and the source-follower amplifier. In this case, the system, transducer, cable, and electronics
function as a tuned circuit and must be calibrated as a system. This may also be the case today when
separate, miniature, solid-state, source-follower-type voltage amplifiers are used in line with the
transducer for signal conditioning.
Because of their high-voltage, low-impedance output (<100 ohms), IEPE transducers provide more
immunity to the cable problems previously listed , especially when they are used with long cables in field,
underwater, or industrial environments. However, some IEPE sensors incorporating JFET inputs have a
higher output impedance (approximately 1500 ohms) and do require more attention to the type and length
of cable used. It is good practice, with any long cable run and with any type of transducer, to use signal
insertion in order to calibrate and to understand the effect long cables on the amplitude, frequency, and
system operation.
8.3

Voltage amplifier

In the 1950s and early 1960s, before the advent of the charge amplifier, voltage amplifiers were
commonly used to condition the high-impedance output from piezoelectric transducer into a lowimpedance voltage signal compatible with recording instruments.
Voltage amplifier functions include isolation, impedance-matching and/or amplification. The isolation
voltage amplifier is used to isolate the input common from the output common of the amplifier. This type
of amplifier is utilized to eliminate ground-loop problems where the transducer cannot be isolated from
ground. The amplifier used for impedance-matching has a high-input impedance, and a low-output
impedance. The high-input impedance reduces the loading effects of the amplifier on the transducer.
Special purpose impedance-matching amplifiers are frequently used with piezoelectric transducers.
These amplifiers have extremely high-input impedances to minimize the unavoidable effects of lowfrequency roll off due to a short time constant.
The voltage amplifier also provides amplification of the transducer signal to a level that is desirable for
good recording. Often, the amplifier can provide a gain factor that will make the output voltage proportion
to input pressure, thus presenting a reading in engineering units. AC-coupled voltage amplifiers exhibit
poor low-frequency response, whereas DC amplifiers have low-frequency capability for static and
dynamic pressure measurement.
8.3.1

Slew rate

For large signals the amplifier rise time is limited by the slewing rate, which is the maximum rate of
change of output voltage of which the amplifier is capable. It should not be confused with amplifier rise
time and bandwidth specifications, which usually apply only for small signals. Slew rate limits the
maximum frequency and minimum rise time for which the amplifier can produce a full-power response. If
the input signal is sinusoidal, the maximum slope of the output voltage should be
(Eq. 8.2)

de
= 2 fE
dt

When this slope exceeds the amplifier slew rate, the output will be distorted (References 40 and 77).
8.3.2

Advantages of the voltage amplifier

The voltage amplifier was, at one time, perhaps the easiest and simplest to use. Due to its simplicity, it
was the lowest in cost of the signal-conditioning amplifiers. The integrated circuit has reduced the cost
and size such that the amplifier can now be located inside or close to the transducer, thus reducing the
attenuation effects of the cable on the transducer sensitivity. Of all the signal-conditioning techniques, the
voltage amplifier has the highest frequency response capability.

ISA-37.16.01-2002

8.3.3

- 46 -

Disadvantages of the voltage amplifier

Voltage amplifiers with long-time constants are difficult to obtain. Further, when used separately with
piezoelectric transducers, the amplifier input voltage is dependent on the total cable capacitance and thus
on cable length.
8.4

Charge amplifier

The charge amplifier (Reference 70) converts the output charge of piezoelectric transducers to a voltage
signal. Most charge amplifiers consist of two sections: a charge converter and a voltage amplifier section.
The amplification section is frequently adjustable so that the transducer can produce a convenient scaled
output.
8.4.1

Frequency response of the charge amplifier

In order to achieve stable operation, most charge amplifiers make use of a feedback resistor connected
between the input and the output of the amplifier section in parallel with the feedback capacitor. The time
constant of the system is then defined by the RC time constant of feedback resistance and feedback
capacitance. Often several feedback resistors with switches are provided, allowing the operator to select
the desired time constant by switching the proper resistor into the circuit.
The laboratory charge amplifier, invented in the early 1950s, did not come into popular use until the
1960s. Early models incorporated high-input impedance electrometer tubes, which were replaced in the
mid-1960s by Field Effect Transistors (FETs). Over the years, five general types of charge amplifiers have
evolved. These include laboratory models, miniature airborne, solid state "in-line" types, and
microelectronic chip designs, which are incorporated directly inside integral electronic piezoelectric
transducers.
1) The "electrostatic" laboratory-type charge amplifier, with a MOSFET input, has very long input time
constants to permit static calibration and short-term static measurements with quartz pressure and
force transducers. It usually has selectable long-, medium- and short-input time constants, multirange capability, and the ability to handle large charge inputs. Single-range versions with minimum
operating controls are often used in dedicated applications, such as signal conditioning for calibration
standards and production test.
2) The laboratory "vibration-type charge amplifier, with JFET input, is designed for lowest noise
operation, mainly for use with piezoelectric accelerometers. They generally operate down to about .2
Hz and do not have the long-discharge, time-constant characteristic of the electrostatic type. Early
models had limited frequency response to 20 KHz, however, most current models operate to 100
KHz.
3) The airborne-charge amplifier, packaged in a small rugged housing, generally has input sensitivity
adjust, bias adjust, and capability for limiting and filtering the output for telemetry transmission or the
measurand.
4) There is also a category of miniature, single-range, solid-state charge amplifiers. They are intended
for mounting close to the sensor in long cable-driving applications to provide improved signal to noise.
They generally operate from a constant-current power source similar to IEPE sensors. This
installation provides a better signal to noise ratio than would be attainable with long input cables
feeding into a laboratory-type charge amplifier.
5) Microelectronic hybrid-charge amplifiers are commonly used inside integral electronic PE transducers
incorporating ceramic-sensing elements. The circuits are usually encapsulated to withstand high
shocks. Depending on whether a particular transducer has a JFET or MOSFET input stage
(Reference 69), performance characteristics such as noise, dynamic range, cable-driving capability

- 47 -

and operating temperature range may vary. IEPE transducers incorporating quartz elements most
often use a simplified unity gain source follower circuit.
If no feedback resistor is used, the time constant of the system will be equal to the product of feedback
capacitor leakage resistance to ground and amplifier gain. Since the inner loop gain of the charge
amplifier is normally very high, in the order of 10,000 or more, very long time constants can theoretically
be obtained with such an arrangement, even if the cable and the transducer offer relatively low insulation
values. However, such a system will generally be unstable and, if not perfectly balanced, will keep drifting
until it biases itself into cutoff position.
The high-frequency response of the charge amplifier is strongly influenced by the feedback ratio, the ratio
between the feedback capacitor used and the capacitive loading at the input multiplied by the gain of the
amplifier. When the amplifier is being operated on a sensitive range, and long input cables with large
shunt capacitance are used, a severe degradation of the frequency response of the charge amplifier may
result. Different range settings will result in quite different values for the high-frequency cutoff of the
charge amplifier. A high-frequency cutoff filter is often used in order to achieve a more uniform frequency
response over the range of available gain positions. As with the voltage amplifier, the charge amplifier
slew rate is the maximum rate at which the output voltage can change from one level to another. To
prevent distortion in the output, the amplifier slew rate should not be exceeded.
8.4.2

Advantages of the charge amplifier

The main advantage of the charge amplifier is that the system sensitivity is unaffected by the length of
cable between the transducer and the charge amplifier. This is advantageous in that the transducer can
be calibrated with one length of cable, and the calibration will be valid for any other length of cable in a
particular application. Another capability of the electrostatic charge amplifier is its long discharge time
constant, which provides for static calibration of quartz pressure and force transducers and allows
measurement of short-term static events.
8.4.3

Disadvantages of the charge amplifier

The noise level increases with an increase in cable length. With long cables, the charge amplifier is
particularly sensitive to spurious signals generated in the cable. Although the signal level is unaffected,
the noise increase with long cable length results in a low signal-to-noise ratio for the system. The
vibration-type charge amplifier with JFET input is less affected by these conditions than the electrostatic
type. Since the piezo-transducer/low-noise cable/charge amplifier operates at very high impedance, they
must be kept clean and moisture free. Connections at the transducer must be sealed.

9
9.1

Data acquisition and analysis

Digital oscilloscope or recorder

This instrument is probably used more than any other for measuring dynamic output of sensors. It is a
universal instrument of relatively low cost compared to other instruments with equivalent capability. When
using this instrument as part of a dynamic-pressure calibration system, particular attention must be paid
to the frequency response and rise-time characteristics.
Rise-time characteristics are of importance when using pulse inputs or shock-tube methods for sensor
calibration. The rise time is generally a good indication of relative bandwidth. For reasonably accurate
rise-time measurements the oscilloscope should respond at least five times faster than the signal to be
measured. The rise time of any scope and preamplifier must be established before dynamic-pressure
calibrations are made. This rise time is measured by application of a square wave to a properly
terminated oscilloscope input.

ISA-37.16.01-2002

9.2

- 48 -

Data analysis

The data analysis to be performed depends on the goal of the activity (e.g., amplitude calibration,
transfer-function determination, etc.). Each of these goals has its own individual considerations
associated with it.
9.2.1

Time analysis

Analysis in the time domain only requires the faithful reproduction of the waveform of the test and
reference transducer. The frequency response that is provided to each data channel should be the
minimum of 1/(reciprocal of the pulse duration) or 0.35/(pulse rise time). Both a flat amplitude response
and linear phase integrity should be maintained over this frequency range. The digital sampling rate, in
order to satisfy the Nyquist criteria, should be placed at 1/(2x90 percent attenuation frequency of
measuring system). This assures no aliasing of the pressure signal due to high-frequency noise in the
recorded signal. The sampling rate of recorders is so high that this should not be a problem. However,
this assumption should be verified.
9.2.2

Frequency analysis

For the analysis of pulse signals, multi-channel spectrum analyzers are readily available. A given channel
can have its Fourier spectrum calculated, or a transfer function (input/output) relationship can be
determined. Coherence functions should be provided if transfer functions are calculated. The coherence
function should be near 1.0 over the frequency range where the transfer function is evaluated.
Shock-tube-based pressure responses provide unique challenges. The signal becomes undefined once
the relief wave attenuates the signal. Specific work has been performed to permit analyses to occur.

10 Reporting test results

The success or failure of a test program may, in the final analysis, depend on the quality of the reported
information. Test data must be presented in a meaningful manner. The amount of background information
required in a test report will depend on the complexity and uniqueness of the tests. Specialized
equipment and techniques require a detailed explanation of the test conditions and complete supporting
information. This clause considers topics directly related to the effective presentation and communication
of dynamic calibration data.
10.1 Test conditions
10.1.1 Test equipment
The quality of calibration information is directly dependent on the equipment and methods used to obtain
it. Since most dynamic-calibration equipment is non-standard or even unique, some definitions of its
capabilities will always be necessary.
Generally, the characteristics of test equipment may be established in several ways:
1) By mathematical calculations supported by experiment.
2) By testing or calibrating the test equipment against known standards.
3) By some combination of 1) and 2).
Test-equipment characteristics must be clearly outlined to the reader. It should be recognized that the
characteristics of the test equipment may vary with the conditions of use. In any test reported, the test
equipment should be used, preferably, under the same conditions at which its characteristics were

- 49 -

established, or justification should provided for use under other conditions. Lack of comment in this area
will leave doubts in the readers mind as to whether the characteristics of the test equipment were
sufficiently well understood and considered.
One technique for communicating information quickly on the characteristics of dynamic pressure
generators is to present the recorded-time histories of the reference transducer output. Similar time
records showing the response of the readout equipment to electronically generated square waves and
sine waves tell much about the readout equipment characteristics. Lastly, the determination of the
transfer function of the measurement system with its associated coherence is extremely useful.
10.1.2 Environmental and related conditions
The rapidly changing pressure required for the dynamic calibration of pressure transducers is often
accompanied by other dynamic environmental conditions such as temperature, vibration, and
acceleration. These conditions may significantly affect the transducer output. Comments or data
explaining to what extent these conditions affect the test results should be presented, such as a recording
photograph showing the output of a transducer whose diaphragm is blanked off from pressure but whose
case is exposed to the same acceleration forces normally present during a shock-tube test.
10.2 Results and discussion
Having discussed the test conditions and any effects they may have on the transducer response
characteristics, the presentation of the test results is fairly straightforward. An illustration of the
transducers response to some type of dynamic pressure input should be included in every report. One of
the most useful illustrations is a display of the transducers response to a shock tube pressure step.
Illustrations of response to dynamic pressure inputs should be clearly labeled in consistent terms. Terms
used in the explanation of transducer response characteristics should be defined.
The primary reason for this arises from the lack of agreement between investigators as well as different
individuals interpretation of transducer response characteristics and their relation to basic frequency
response. If the transducers response characteristics are reported (preferably accompanied by an
illustration), the user can draw his own inferences as to transducer frequency response and judge if the
transducer will be suitable for its intended use.

11 References
1) ISA37.11975 (Reaffirmed 1982), Electrical Transducer Nomenclature and Terminology, ISA
The Instrumentation, Systems, and Automation Society, 1975.
2) Schweppe, J. L., Eichenberger, L. C., Muster, D.F., Michaels, E. L., and Paskusz, G. F., Methods for
the Dynamic Calibration of Pressure Transducers, NBS Monograph 67, December 12, 1963.
3) White, Gifford, Response Characteristics of a Simple Instrument, Statham Instrument Note 2, AprilMay 1948.
4) White, G. E., The Meaning Of Natural Frequency, Statham Instrument Note 12, NovemberDecember 1949.
5) Pennington, D., Piezoelectric Accelerometer Manual, Endevco Corporation, Pasadena, California,
1965.
6) Beckwith, T. G. and Buck, N. L., Mechanical Measurements, Addison Wesley Publishing Co., Inc.,
1961.
7) Norton, H. N., Handbook of Transducer for Electronic Measuring Systems, Prentice-Hall, Inc. 1969.

ISA-37.16.01-2002

- 50 -

8) ISA Transducer Compendium, Second Edition Part 1, ISA The Instrumentation, Systems, and
Automation Society, 1969.
9) Pressure Transducers, Home Study Course 18, Measurements and Data, November-December,
1969.
10) Beranek, L. L., Acoustic Measurements, John Wiley & Sons, Inc. 1949.
11) ISA37.31982 (Reaffirmed 1995), Specifications and Tests for Strain Gage Pressure Transducers,
ISAThe Instrumentation, Systems, and Automation Society, 1982.
12) ISA37.101982 (Reaffirmed 1995), Specifications and Tests for Piezoelectric Pressure and SoundPressure Transducers, ISA The Instrumentation, Systems, and Automation Society, 1982.
13) Smith, R. O., A Liquid Medium Step-Function Pressure Calibrator, American Society of Mechanical
Engineers, Journal of Basic Engineering, December 1965 (ASME Paper No. 63-WA263).
14) Rathke, T. S., A System for the Dynamic Calibration of Pressure Transducers," Paper presented at
nd
the Instrument Society of America 22 Annual ISA Conference and Exhibit, September 11-14, 1967,
Chicago, Illinois (Preprint Number P19-2PHYMMID-67).
15) Rockwell, D. W., Dynamic Calibration of Pressure Transducers, Metrology Engineering Center,
Navy, Engineering Circular 15, 1967.
16) Aronson, P. M., and Waser, R. H., Pressure-Pulse Generator for the Calibration of Pressure Gages,
Naval Ordinance Laboratory Technical Report 63-143, August 1963.
17) Fenrick, W. J., and Sanders, B. R., Studies on Shock Wave Pressure Gages, VIII, The Pressure
Pulsing Unit (U), Suffield Technical Note No. 125, Suffield Experiment Station, Ralston, Alberta.
18) Morefield, R. I., Dynamic Comparison Calibrator for High Range Pressure Transducers, paper
presented at Institute of Environmental Sciences 1965 Annual Technical Meeting Proceedings.
th

19) Simpson, J. P., and Gatley, W. L., Dynamic Calibration of Pressure Measuring Systems, 16
Aerospace Instrumentation Symposium of ISA, May 1970.

20) Nyland, T. W., Sinusoidal Pressure Generator for Testing Pressure Probes, Advances in
th
Instruments, Vol. 25, Part 2, 621-70, Proceedings of the 25 Annual ISA Conference, (also as NASA
Technical Memorandum NASA TM X-52873).
21) Kistler Instrument Co. Data Sheet 210-2/671.
22) Layton, J. P., and Thomas, J. P., Final Summary Technical Report on Transient Pressure Measuring
Methods Research, Princeton University Aeronautical Engineering Report No. 595t, Princeton
University, Princeton, New Jersey, March 1967.
23) Robinson, R. E., and Liv, C. Y., Resonant Systems for Dynamic Transducer Evaluations, Battelle
Memorial Institute, Columbus Laboratories, Report No. NASA CR-72535, August 31, 1968.
24) Liepmann and Roshko, Elements of Gas Dynamics, John Wiley and Sons, Inc. 1965.
25) Shapiro, A. H., The Dynamics and Thermodynamics of Compressible Fluid Flow, Vol. I, The Ronald
Press Co. 1953.

- 51 -

26) Wittliff, C. E., Wilson, M. R., and Hertzberg, A., The Tailored Interface Hypersonic Shock Tunnel,
paper presented at the SAME-ARS Aviation Confererence, Dallas, Texas, March 16-20, 1968. See
also Journal of the AeroSpace Sciences, Vol. 26, No. 4, April 1959.
27) Lederer, P. W., Methods for Performance Testing of Electromechanical Pressure Transducers, NBS
Technical Note 411, February 1967.
28) Glass, I. I., and Hall, J. G., Shock Tubes, Section 18, Handbook of Supersonic Aerodynamics,
NAVORD Report 1488 (Vol. 6), Bureau of Naval Weapons, (available from the U.S. Government
Printing Office, Washington, D.C.) December 1959.
29) Bentley, W. C., and Walter, J. J., Dynamic Response Testing of Transient Pressure Transducers for
Liquid Propellant Rocket Combustion Chambers, Princeton University, May 1963.
30) Nagar, R., and Ikegami, M., Errors of an Indicator Due to a Connecting Passage, Bulletin JSME,
Vol. 8, No. 29, 1965.
31) Horn, L., The Response of Flush Diaphragm Pressure Transducers to Thermal Gradients, Preprint
th
No 13.3-4-65, 20 Annual ISA Conference Proceedings Vol. 20, 1965, Part 3.
32) Rogero, S., Measurement of the High Frequency Pressure Phenomena Associated with Oscillatory
rd
Combustion in Rocket Engines, Proceedings of the ICRPG, 3 Combustion Conference, JFK Space
Center, October 1966.
th

33) Rogero, S., Internal Rocket-Engine Probes for Use in a Combustion Instability Environment, 15
National Aerospace Instrumentation Symposium, May 1969.

34) Special Considerations for Combustion Instability Instrumentation and Data Presentation, prepared
by the Committee on Instrumentation and Test Data of the Interagency Chemical Rocket Propulsion
Group (ICRPG) CPIA Publication No. 170, June 1968.
35) Arlowe, H. D., Electrical Noise in Instrumentation Systems, pages 341-347, Institute of
Environmental Sciences, Proceedings, 1963.
36) Flynn, George, Special Report on Signal Conditioning," Electronic Products, May 1968.
37) Judd, John and Morgan, Edward P., Zero-Drive" A New Zero Impedance Signal Conditioning
Concept for Piezo-Electric Transducers, Instrument Society of America paper, P13-2-PHYMMID-67,
nd
presented at 22 Annual ISA conference and Exhibit, Chicago, Illinois, September 11, 1967.
38) Kuo, Franklin F., Network Analysis and Synthesis, John Wiley and Sons, Inc. New York, 1962.
39) Morrison, R., Survey of Instrumentation Amplifiers, Electronic Instrument Digest, March 1967.
40) Stein, Peter K., Measurement Engineering, Vol. I, Basic Principles, Stein Engineering Services, Inc.,
1964.
41) Morrison, R., Amplifier Slewing Rate, Measurements Data, September-October 1967.
42) Bouche, R. R. and Ensor, L. C., Use of Reciprocity Calibrated Accelerometer Standards for
Performing Routine Laboratory Comparison Calibrations, Endevco Corporation TP 226, October
1964.
43) Pogach, B., Transducer Conditioning 1967, Preprint Number 16.17-66, 21st Annual ISA Conference
1966.

ISA-37.16.01-2002

- 52 -

44) Accelerometer Calibration, Weston-Lexington Application Notes AN 1-68/PR840/008, Weston

Instruments, Inc.
45) Harris, C. M. and Crede, C. E., Shock and Vibration Handbook, Volume I, McGraw-Hill Book Co.,
Inc., New York 1961.
46) Majewski, E. H., Determining Second-Order, Response-Underdamped Case, Control Engineering,
June 1969.
47) Daniel, R. G., Experimental Frequency Response Determination, Instruments and Control System,
June 1969.
48) Laspe, C. G., Determining Frequency Response From Transient Response, Instrument and Control
Systems, December 1964.
49) Favour, J. D., and Steward, R., Primary Calibration of Pressure Transducers to 10,000 Hz, 15
Aerospace Instrumentation Symposium of ISA, May 1969.

th

50) Darsey, D. M., Calibration of Systems by Dynamic Response Analysis, Advances in

th
Instrumentation, Vol. 25, Part 2, 651-70, Proceedings of the 25 Annual ISA Conference, 1970.
51) The Modal Shop, Inc. A PCB Group Company Product Data Sheet Precision Acoustical Calibration
Work Station Model 9350C.
52) GRAS Sound & Vibration Product Data Sheet Pistonphone Model 42AA.
53) Larson Davis Inc. A PCB Group Company Product Data Sheet Acoustical Calibrator Model Cal.
200.
54) Lally, R.W., "Application in Integrated Circuits to Piezoelectric Transducers," PCB Piezotronics, Inc.,
ISA, PrePrint No. P4-2-PHYMMID-67.
th

55) Granath, Ben, "New Dynamic Pressure Generator and its Applications," PCB Piezotronics, 18
Transducer Workshop, Colorado Springs, June 20-22, 1995.

56) Bean, V.E. and Rosasco, G.J., NIST "Workshop on the Measurement of Transient Pressure and
Temperature," Proceedings, NISTIR-4828, April 23-24, 1991.
57) Lally, J.F., "Dynamic Pressure-Calibration Instruments and Sensor-Transient Response," Summary,
PCB Piezotronics, Inc., Paper presented at Seventh Transducer Workshop, San Diego, CA, June 2224, 1993.
58) Coulter, G.A., "Dynamic Calibration of Pressure Transducers at BRL Shock-Tube Facility,"
Memorandum Rept. No. 1843, Aberdeen Proving Ground, MD, May 1967.
58) Lally, J.F., "Aronson Shockless Pressure Step Generator," PCB Piezotronics, Inc., Paper presented
at the Thirteenth Transducer Workshop, Monterey, CA, 1985.
59) "Accelerometer Cable, Wiring, and Connections," Recommended Practices, PCB Piezotronics, Inc.,
Technical Bulletin No. 59.
60) Juhasz, A.A., Bullock, C.D., and Newhall, D.H., "Calibration of Ballistic Pressure Transducers,"
Measurement and Control Magazine, April 1991.

- 53 -

61) Riegebauer, J., "Dynamic Pressure Calibrator (Drop Hammer Method), AVL List GmbH, Graz,
Austria, presented at the NIST Workshop on Measurement of Transient Pressure, 1991.
62) PCB Piezotronics, Inc., Pressure Catalog, PFS-100A, pages 49-55, November, 1998.
63) Harris, C. M., Editor, Shock and Vibration Handbook, Third Edition, McGraw - Hill Book Company,
1993.
64) "Design of a Cold Gas Shock Tube," Texas Christian University Engineering Class, Fort Worth, TX,
1997.
65) Resch, Gerhard, "Dynamic Conformity and Dynamic Peak Pressure Accuracy - Two New Features of
th
Pressure Transducers," 14 Transducer Workshop, Colorado Springs, CO (available Secretariat
Range Commanders Council, White Sands Missile Range, NM), June, 1987.
66) Hilten, J.S., Lederer, P.S., Vezzetti, C.F., and Mayo-Wells, J.F., "Development of Dynamic
Calibration Methods for Pogo Pressure Transducers", NBS (now NIST) Technical Note 927,
November 1976.
67) Sachs, D.C., Cole, E., "Air Blast Measurement Technology," Defense Nuclear Agency Report DNA
4115F, September 1976.
68) Petersen, C., Coleman, P., "Small-Diameter Bar Gauges for Fast-Response Air-Blast
th
Measurements," 17 Transducer Workshop, San Diego, CA (available Secretariat Range
Commanders Council, White Sands Missile Range, NM), June 1993.
69) Dally, J.W., Riley, W.F., McConnell, K.G., Instrumentation for Engineering Measurements, John
nd
Wiley and Sons, 2 Edition, 311-317, 1993.
70) Vezzetti, C.F., Hilten, J.S., Mayo-Wells, J.F., Lederer, P.S., "A New Dynamic Pressure Source for the
Calibration of Pressure Transducers," NBS (now NIST) Technical Note 914, June 1976.
71) Harris, C. M., Editor, Shock and Vibration Handbook, Third Edition, McGraw-Hill Book Company,
1993, pp. 15-15 to 15-16.
th

72) Granath, B., "New Dynamic Pressure Generator and Its Application," 18 Transducer Workshop,
Colorado Springs, CO (available Secretariat Range Commander Council, White Sands Missile
Range, NM), June 1995.
th

73) Kubler, J.M., Roesli, V., "Ballistics Pressure Transducers and Their Applications," 14 Transducer
Workshop, Colorado Springs, CO (available Secretariat Range Commanders Council, White Sands
Missile Range, NM), June 1987.
74) Walter, P.L., "Dynamic Force, Pressure, and Acceleration Measurement," Measurement and Control,
Issue 194, (Part 1 of 12 consecutive parts published by the author bi-monthly, 90-96), April 1999.
75) Morrison, R., Grounding and Shielding Techniques in Instrumentation, Third Edition, John Wiley and
Sons, 1986.
76) Ott, H.W., Noise Reduction Techniques in Electronic Systems, Second Edition, John Wiley and Sons,
1988.
77) Sedra, A.S. and Smith, K.C., Microelectronic Circuits, Fourth Edition, Oxford University Press, 1997.

This page intentionally left blank.

Developing and promulgating sound consensus standards, recommended practices, and technical
reports is one of ISAs primary goals. To achieve this goal, the Standards and Practices Department
relies on the technical expertise and efforts of volunteer committee members, chairmen, and reviewers.
ISA is an American National Standards Institute (ANSI) accredited organization. ISA administers United
States Technical Advisory Groups (USTAGs) and provides secretariat support for International
Electrotechnical Commission (IEC) and International Organization for Standardization (ISO) committees
that develop process measurement and control standards. To obtain additional information on the
Societys standards program, please write:
ISA
Attn: Standards Department
67 Alexander Drive
P.O. Box 12277
Research Triangle Park, NC 27709
ISBN: