ANSI/NCSL Z540.3:2006

Transcription

December 2006
measure
NCSL INTERNATIONAL
The Journal of Measurement Science
Vol. 1 No. 4 • December 2006
NCSL International
In This Issue:
measure • The Journal of Measurement Science
Uncertainties Related to Thermal
Expansion in Dimensional Metrology
Gravimetric Calibration of
Volumetric Standards with
Capacities Exceeding Five Gallons
A Theory for RF and Microwave
Scalar Reflectometer Errors
ANSI/NCSL Z540.3:2006:
Requirements for the Calibration
of Measuring and Test Equipment
Vol. 1 No. 4
measure
N C S L I N T E R N AT I O N A L
T h e J o u r n a l o f M e a s u re m e n t S c i e n c e
WELCOME to NCSLI measure,
a metrology journal published by
NCSL International (NCSLI), for the
benefit of its membership.
Contents
Features
22
2 0 0 7 N C S L I Wo r k s h o p
& Sy mpos ium
26
See Page 19
32
S P E C I A L R E P O RTS
T he C I P M Wo r k i n g G ro u p o n M e t ro l o g y o f M a t e r i a l s
Seton Bennett and Graham Sims
AN S I/ N C SL Z 5 40 . 3: 2 0 06 : Re qu ir e m e n t s f o r t h e C a l i b r a t i o n
o f M e a s u r i n g a nd Te s t E q u i p m e n t
Del Caldwell
T E C H N I C A L PA P ER S
Unc er tainties Re late d to T her mal Exp ans ion
i n D i m e n s i o n a l M e t ro l o g y
Ted Doiron
38
A T h e o r y f o r R F a n d M i c ro w a v e S c a l a r R e f l e c t o m e t e r E r r o r s
Robert D. Moyer
46
A D i re c t C o m p a r i s o n S y st e m f o r M e a su r i n g
R a d i o F re q ue nc y P o we r ( 100 kH z to 18 G Hz )
Ronald Ginley
50
Remo te Tim e C alibratio ns
vi a t h e N IS T T i m e M e a s ure m e n t a n d A n a l y si s S e r v ic e
Michael A. Lombardi and Andrew N. Novick
60
R E V I E W PA P ER S
G r a v i m e t r i c C a l i b r a t i o n o f Vo l u m e t r i c S t a n d a rd s
w it h Cap ac it i es E x c eed in g F i v e G al l o ns
L.F. Eason
Departments
CONTACT NCSLI
3
L e t t e r f ro m t h e E d i t o r
B u s i n e s s O ff i c e :
Craig Gulka, Business Manager
NCSL International
2995 Wilderness Place, Suite 107
Boulder, CO 80301-5404 USA
Phone: 303-440-3339
Fax: 303-440-3384
Email: [email protected]
3
Le tt ers
5
I n t e r na t i o n al NM I Ne w s
12
M e t ro l o g y N e w s
75
N e w P ro d u c t s
79
A dv e rt i s er In d e x
80
Cla ssifie ds
NCSLI measure I n f o r m a t i o n :
www.ncsli.org/measure/
MEASURE
|
1
NCSLI Member Benefit
in the Spotlight
P u b l i c a t i o n s a n d Vi d e o s
NCSL International has developed an extensive library of
publications and training videos to educate NCSL International
members about various areas in metrology, including:
International standards; laboratory procedures; measurement
practices; and current metrology services and seminars. Some of
the publications available through NCSL International include:
• ANS/ISO/IEC 17025:2005
• ANSI/NCSL Z540-1-1994 (R2002) Standard
• ANSI/NCSL Z540-1-1994 Handbook
• ANSI/NCSL Z540-2-1997 (R2002) "U.S. Guide to the
Expression of Uncertainty in Measurement"
• Recommended Practices (RPs)
Calibration Intervals; Laboratory Design; Laboratory
Environments; Interlaboratory Comparisons; Determining
and Reporting Measurement Uncertainties; Calibration
Procedures; and more.
• Recommended Intrinsic/Derived Standards Practices (RISPs)
Array Josephson Junction; Quantized Hall Resistance;
Deadweight Pressure Gauges; and Two-Pressure, TwoTemperature Humidity Generator.
• Laboratory Management
Acronyms and Abbreviations; Glossary of MetrologyRelated Terms; Guide to Achieving Laboratory
Accreditation;
Calibration
Laboratory
Manager's
Handbook;
Comparison
Between
ANSI/ISO/IEC
17025:2000 and 17025:2005; Benchmarking Survey
Results; and more.
measure
N C S L I N T E RN AT I O N A L
NCSLI measure (ISSN #19315775) is a metrology journal published by
NCSL International (NCSLI). The journal's primary audience is
calibration laboratory personnel, from laboratory managers to project
leaders to technicians. measure provides NCSLI members with
practical and up-to-date information on calibration techniques,
uncertainty analysis, measurement standards, laboratory accreditation,
and quality processes, as well as providing timely metrology review
articles. Each issue will contain technically reviewed metrology articles,
new products/services from NCSLI member organizations, technical
tips, national metrology institute news, and other metrology information.
Information for potential authors, including paper format, copyright
form, and a description of the review process is available at
www.ncsli.org/measure/ami.cfm. Information on contributing Technical
Tips, new product/service submission, and letters to the editor is
available at www.ncsli.org/measure/tc.cfm. Advertising information is
available at www.ncsli.org/measure/ads.cfm.
Managing Editor
Richard B. Pettit, Sandia National Laboratories (Retired), 7808 Hendrix,
NE, Albuquerque, NM 87110 USA. Email: [email protected]
NMI/Metrology News Editor:
Michael Lombardi, NIST, Mailcode 847.00, 325 Broadway, Boulder, CO
80305-3328 USA. Email: [email protected]
New Product/Service Announcements:
NCSLI Business Office, 2995 Wilderness Place, Suite 107, Boulder, CO
80301-5404 USA. Email: [email protected]
Technical Support Team:
Norman Belecki, Retired, 7413 Mill Run Dr., Derwood, MD 20855-1156.
Belinda Collins, National Institute of Standards and Technology (NIST),
USA
Salvador Echeverria, Centro Nacional de Metrologia (CENAM), Mexico
Andy Henson, National Physical Laboratory (NPL), United Kingdom
Klaus Jaeger, Jaeger Enterprises, USA
Dianne Lalla-Rodrigues, Antigua and Barbuda Bureau of Standards,
Antigua and Barbuda
Angela Samuel, National Measurement Institute (NMI), Australia
Klaus-Deter Sommer, Landesamt fuer Mess und Eichwesen
Thueringen (LMET), Germany
• Organization listed in the Directory of Standards Laboratories
(Web Site)
Alan Steele, National Research Council (NRC), Canada
• Post Job Listings in the NCSLI Jobs database
Andrew Wallard, Bureau International des Poids et Mesures (BIPM),
France
Pete Unger, American Association for Laboratory Accreditation (A2LA),
USA
• Training Information Directory
Tom Wunsch, Sandia National Laboratories (SNL), USA
• NCSL International MEASURE Journal and Quarterly
Newsletter
Production Editor:
Mary Sweet, Sweet Design, Boulder, CO 80304 USA
• Training Video Tapes
For more information about membership benefits, contact the
NCSLI Business Office at 2995 Wilderness Place, Suite 107,
Boulder, CO 80301 USA
(Phone: 303-440-3339)
or visit the web site at
www.ncsli.org/memberships/
2
|
MEASURE
Copyright © 2006, NCSL International. Permission to quote excerpts or to reprint
any figures or tables should be obtained directly from an author. NCSL
International, for its part, hereby grants permission to quote excerpts and reprint
figures and/or tables from this journal with acknowledgment of the source.
Individual teachers, students, researchers, and libraries in nonprofit institutions and
acting for them are permitted to make hard copies of articles for use in teaching or
research, provided such copies are not sold. Copying of articles for sale by
document delivery services or suppliers, or beyond the free copying allowed above,
is not permitted. Reproduction in a reprint collection, or for advertising or
promotional purposes, or republication in any form requires permission of one of
the authors and written permission from NCSL International.
www.ncsli.org
Letter from the Editor
Letters
This is the fourth and final issue of NCSLI measure for
2006. It has been a very enjoyable and educational experience for me, and I have learned a lot about the electronic
publishing business. Each issue has surpassed my high
expectations, and I am very pleased with the positive
response we have received from the NCSLI membership.
Again, I would like to thank the NCSLI Board of Directors,
and especially Tom Wunsch, of Sandia National Laboratories, who conceived of the journal format and has continued
to provide ideas and leadership for each issue.
As I think about 2007, my vision is to expand the journal from the current 80
pages to 96 pages. Most of that expansion will include more Technical Papers, as
well as Special Reports, Review Papers, and Technical Tips. Currently we are
averaging about eight papers per issue. With the expanded journal, we will be
able to publish about ten papers per issue. In addition, we will have more space
for NCSLI member organization new product/service announcements and for
paid advertisements.
Recently, measure has received several technical papers from international
sources, including Canada, Egypt, Japan, and Chile. This is strong evidence that
our journal is both being recognized as a valuable publication resource in metrology, and that it is obtaining international recognition. As part of my NCSLI
involvement as Vice-President of Measurement Science and Technology, I have
been working with Jim Wheeler, of the U.S. Navy in his role as the chair of the
Measurement Comparisons Programs Committee to request that every NCSLI
sponsored interlaboratory comparison (ILC) results in an article in measure. In
this way, the important results and lessons learned from each NCSLI ILC will
become part of our permanent record and will better benefit all NCSLI members.
I also want to acknowledge a special thanks to Michael Lombardi, NIST, who
has collected and edited the NMI and Metrology News items. Mike has also contributed several excellent technical papers and review articles documenting the
important services of the Time and Frequency Division at NIST, Boulder. He has
also provided very valuable editorial suggestions and assistance as we strive to
develop our own style and processes for the journal. Mike has a strong interest
and educational background in technical writing – and they have served us well!
His ideas and support are greatly appreciated.
Finally, the NCSLI Business Office is in the process of developing an online
database with information on each paper that is submitted for publication, the
long list of technical reviewers who provide the valuable service of reading and
commenting on each paper, new product/service announcements, NCSLI member
advertisers, etc. When completed, this new tool will streamline the multitude of
tasks that need to be completed in order to publish each edition.
Thank you again for your support and interest in measure.
Thaks for telling your readers about
"The Measurement Blues." I hope they
enjoyed listening to it as much as I
enjoyed writing it. I've been kicking
around the idea of writing a blues about
measurements for five years. I didn't
tell anyone about it until I introduced it
to the Test & Measurement World staff
this year. Of course, I had to get the calibration part in. As we all know, measurements are meaningless without
calibration.
Martin Rowe
Senior Technical Editor
Test & Measurement World
Richard Pettit
Managing Editor
Sandia National Laboratories (Retired)
H OW TO R EA C H U S: MA I L letters to: NCSLI
Te c h n i c a l P a p e r A m e n d m e n t
It has been brought to my attention that
I inaccurately attributed the development of the first Josephson voltage
system (JVS) in my article “Experimental Design of NCSLI 2005 Josephson
Voltage Standard Interlaboratory Comparison,” published in NCSLI measure,
vol. 1, no. 1, pp. 36-40, March 2006.
The first demonstration of a 1 V JVS
was performed in the framework of
cooperation between Physikalisch-Technische Bundesanstalt (PTB) and NIST. I
should have clearly noted this collaboration and should have referenced the
following publications:
[1] J. Niemeyer, J.H. Hinken, and R.L Kautz,
“Microwave-induced constant voltage
steps at one Volt from a series array of
Josephson junctions,” Appl. Phys. Lett.,
vol. 45, pp. 478-480, 1984.
[2] J. Niemeyer, L. Grimm, W. Meier, J.H.
Hinken, and E. Vollmer, “Stable Josephson reference voltages between 0.1 and 1.3
V for high-precision voltage standards,”
Appl. Phys. Lett., vol. 47, pp. 1222-1223,
1985.
On behalf of myself and my coauthors, I
apologize for my omission.
Dr. Yi-hua Tang
Quantum Electrical Metrology Division
National Institute of Standards
and Technology (NIST)
measure Journal, 2995 Wilderness Pl., Ste 107, Boulder, CO
80301-5404 USA
FA X letters to: 303-440-3384 E - M A I L letters to: [email protected]
MEASURE
|
3
NMI NEWS
NMI NEWS
Qualified NMIs Can Now Display
the CIPM MRA Logo
on Calibration Certificates
It is now possible for National Metrology Institutes (NMIs) to display a new
logo on their calibration and measurement certificates. This logo, designated as the CIPM MRA logo,
consists of the image of the Pavillon de Breteuil designed in blue
and surrounded by a yellow arc. This CIPM MRA acronym
combines the acronym CIPM, the International Committee for
Weights and Measures, and MRA, the acronym commonly used
for Mutual Recognition Arrangement. The National Physical
Laboratory (NPL) in the United Kingdom was the first NMI to
apply to use the new logo.
The purpose of the voluntary use of the CIPM MRA logo on
calibration certificates is to allow NMIs to draw the attention of
their customers, and other interested parties, to the recognition
by all other signatories of the CIPM MRA of the validity of those
certificates. The logo can only be displayed after a formal application is submitted to the International Bureau of Weights and
Measures (BIPM). The CIPM MRA logo can only be displayed
on those calibration and measurement certificates covered by a
Calibration and Measurement Capability (CMC) document published in the Appendix C section of the BIPM key comparison
database (KCDB).
For more information, www.bipm.fr/en/cipm-mra/logo/
INMS Moves Forward with Quality System
The Institute of National Measurement Standards (INMS),
Canada’s NMI, has made considerable progress with their
quality system since 2005. Photometry and Radiometry Standards, Acoustical Standards, Electrical Power Measurements
(EPM), Dimensional Metrology, and Mass Standards are now
listed in Appendix C of the International Committee for Weights
and Measures (CIPM) MRA. Appendix C lists the quantities for
which calibration and measurement certificates are recognized
by NMIs participating in part two of the arrangement. The MRA
gives users reliable and quantitative information on the comparability of national metrology services and provides the technical basis for wider agreements negotiated for international
trade, commerce and regulatory affairs. The Chemical Metrology, Thermometry, Electrical Standards, Time and Frequency,
and Ionizing Radiation Standards laboratories have all recently
completed their quality systems and had successful internal
audits.
The EPM group went through the accreditation assessment
visit for their quality system in November 2005. The accreditation body, which was the Standards Council of Canada, assembled an assessment team which included measurement experts
from NIST (USA) and NMI (Australia). In response to the
assessment findings, all 35 existing procedures were revised and
15 new procedures were written. The EPM was then accepted
by the Interamerican Metrology System (SIM), the regional
metrology organization for the Americas. The Calibration and
Measurement Capabilities (CMCs) of the EPM laboratory are
now included in the BIPM's Key Comparison Database
(KCDB).
In 2005-2006 INMS has participated in the planning or
implementation of 29 inter-NMI comparisons, including some
nine comparisons under the auspices of SIM, which comprises
the 34 nations of North, Central and South America, as well as
the Caribbean nations. A branch of the National Research
Council (NRC), INMS is located in Ottawa and serves as
Canada’s NMI.
For more information about NRC-INMS: inms-ienm.nrc.cnrc.
gc.ca
The New Dawn of Time:
NPL to Move Time Signal Station
The time signal used to set Britain’s clocks with extreme accuracy is on the move from Rugby, where it has been transmitted
since 1927, to a new home in Anthorn on the west coast of
Cumbria. The signal, often referred to as ‘The time from Rugby,’
will in the future be known as ‘The Time from NPL.’
The National Physical Laboratory (NPL), which has been
responsible for the accurate time signal from Rugby since 1950,
will make the switch in April 2007 following the award of a new
contract to VT Communications. The switchover will take place
following a three-month test period at the beginning of next
year, with the final transfer from Rugby to Anthorn taking place
at the end of March 2007. NPL has reassured most users that
they need take no action to continue receiving the signal.
NPL managing director, Steve McQuillan, said: “Maintaining
accurate time is essential to keeping the modern world
working. Most people only need time to be accurate to within a
few seconds or even minutes, but global navigation systems, the
Continued on page 7
There are three atomic clocks housed at the radio stations. NPL
compared these with NPL master clocks in Teddington by making
use of the GPS system. The group of atomic clocks at NPL keep
the UK's time accurate to within one second in three million years.
MEASURE
|
5
NMI NEWS
internet, email, television, power industry, transportation, and
financial systems are just some of the industries that depend on
very accurate time to operate. We are delighted to be working
with VT Communications on the transfer of the time signal to
Anthorn. While most users check their time against the signal
periodically, a small number of organizations use the signal constantly in their work. We regularly notify those we know who
may be affected by our testing and we’ll be happy to add any
other users to our email list if they get in touch. However the
vast majority of time signal users will not experience any disruption during the testing and switchover.”
Doug Umbers, Managing Director of VT Communications,
said: “We are very proud to be working in partnership with NPL
on a program of national significance. We are excited to be
implementing a highly resilient solution, which will provide tangible benefits to all stakeholders.”
The time signal is accurate to within one millisecond of Universal Time and supports a wide range of services, including
emergency 999 communications, train companies, cash
machines, and mobile phone billing systems. These users will
not be affected by the change.
The signal’s transmission is linked to NPL’s atomic clocks at
Teddington in South West London. NPL is home to UK’s atomic
time and one of only five laboratories worldwide using the latest
cesium fountain technology to contribute to the world time
standard Coordinated Universal Time (UTC).
Contact: Joe Bennett, NPL, [email protected]
NIST Noise Measurement May Boost
Cell Phone Performance
Researchers at the National Institute of Standards and Technology (NIST) and industry collaborators have developed
improved methods for accurately measuring very faint thermal
“noise” in electronic circuits, which is caused by the random
motion of electrons. The technique may help improve the signal
range, data rate and battery life of cell phones and other wireless communications devices.
Low background noise typically translates to better performance in electronics, such as longer ranges and clearer signals or
higher information-carrying capacity. However, noise too low to
measure means that circuit designers cannot tune the system for
optimal performance. The NIST research focuses on CMOS
(complementary metal oxide semiconductor) transistors, which
are inexpensive and widely used in integrated circuits for wireless devices. Noise levels for CMOS transistors have, until now,
been too low to measure accurately in much of their signal frequency range (1 to 10 GHz), and as a result CMOS circuits may
be poorly matched to wireless transmission systems, resulting in
significant signal loss.
In a collaboration with IBM Semiconductor Research and
Development Center (Essex Junction, VT) and RF Micro
Devices (Scotts Valley, CA), NIST has developed and demonstrated the capability to reliably measure the noise in CMOS
devices before they are cut from silicon wafers and packaged.
This is believed to be the first method for on-wafer noise measurements directly linked to national standards for thermal noise
power. The new measurement methods were described at the
IEEE Radio Frequency Integrated Circuits Symposium in San
Francisco.1
The team also demonstrated the use of “reverse” noise measurements, which is focusing on noise emitted from the input of
the transistor when incoming signals are reflected and scattered,
as a tool for checking overall noise parameters. This method can
improve precision, particularly of the optimal impedance properties needed in transistors to minimize noise. Reverse noise
measurements also may help improve modeling of CMOS transistors.
NIST Researchers Unveil World’s First
Quantum AC Voltage Metrology System
After 10 years of research, the National Institute of Standards
and Technology (NIST) has unveiled the world’s first precision
instrument for directly measuring alternating current (AC) voltages. The instrument is being tested for use in NIST’s lowvoltage AC calibration service, where it is expected to increase
significantly the measurement precision of industrial voltmeters,
spectrum analyzers, amplifiers
and filters.
Described July 14 at the Conference on Precision Electromagnetic Measurements in Turin,
Italy,2 the patented instrument3 is
based on the same ‘Josephson
junction’ technology used in
NIST’s widely used direct current
(DC) voltage standards, offering
high precision based on quantum
physics principles. A Josephson junction consists of two superconducting pieces of metal separated by a thin insulator or
normal metal. When a fixed DC voltage is applied across it, a
junction responds by generating an AC current that oscillates at
a frequency exactly proportional to the applied voltage.
The new instrument uses arrays of junctions to generate AC
pulses in precisely measured voltage units over a range of audio
frequencies. Arbitrary waveforms can be generated at different
voltage levels for different applications. The new standard
would establish an entirely new method for AC voltage metrology. Until now, AC voltage calibrations have been performed
indirectly, by measuring the heat delivered by an instrument to
a resistor, and comparing that measurement to the heat delivered by a known DC voltage. At low voltages (such as 2 milliContinued on page 8
1 J. Randa, T. McKay, S.L. Sweeney, D.K. Walker, L. Wagner, D.R.
Greenberg, J. Tao, and G.A. Rezvani, “Reverse Noise Measurement
and Use in Device Characterization,” Presented June 12, 2006 at the
IEEE Radio Frequency Integrated Circuits Symposium, San Francisco,
CA.
2 S.P. Benz, C.J. Burroughs, P.D. Dresselhaus, T.E. Lipe and J.R. Kinard,
“100 mv AC-DC transfer standard measurements with a pulse-driven
AC Josephson voltage standard,” Presented at Conference on Precision
Electromagnetic Measurements, July 2006, Turin, Italy.
3 S.P. Benz, C.J. Burroughs, C.A. Hamilton, and T.E. Harvey. U.S. Patent
6,236,344 (issued 5/22/01) “AC And DC Bipolar Voltage Source
Using Quantized Pulses.”
MEASURE
|
7
NMI NEWS
volts), the new AC Josephson junction voltage standard should
improve measurement accuracy as much as 1,000-fold.
The concept for the new device was co-invented by
researchers at NIST and Northrop-Grumman in the mid1990s.4 A number of innovations since then have led to the first
practical system. For instance, to increase the output voltage,
NIST developed “nano-stacked” arrays of Josephson junctions,
in which the spacing between junctions is reduced to less than
100 nanometers by stacking the junctions on top of each other.
Using this technique, NIST can make programmable voltage
standard integrated circuits with over 130,000 junctions on a
single chip. The new AC instrument currently has a maximum
output of 0.10 volts; NIST researchers hope eventually to
increase that level to 1 volt.
For more information, contact Sam Benz, NIST:
[email protected]
Einstein’s Magnetic Effect
Is Measured on Microscale
A gyromagnetic effect, the rotation of an object caused by a
change in magnetization discovered by Albert Einstein and
Dutch physicist Wander Johannes de Haas, has been measured
at micrometer-scale dimensions for the first time at the National
Institute of Standards and Technology (NIST). The new
method may be useful in the development and optimization of
thin film materials for read heads, memories and recording
media for magnetic data storage and spintronics, an emerging
technology that relies on the spin of electrons instead of their
charge as in conventional electronics.
The Einstein-de Haas effect was first observed in experiments
reported in 1915, in which a large iron cylinder suspended by
a glass wire was made to rotate by an alternating magnetic field
applied along the cylinder’s central axis. By contrast, the NIST
experiments, described in Applied Physics Letters5, measured
the Einstein-de Haas effect in a ferromagnetic thin film only 50
nanometers thick deposited on a microcantilever, which is a tiny
beam anchored at one end and projecting into the air. An alternating magnetic field induced changes in the magnetic state of
the thin film, and the resulting torque bent the cantilever up and
down by just a few nanometers.
Using a laser interferometer to measure the movements of the
cantilever and comparing those data to changes in the magnetic
state of the material, researchers were able to determine the
“magnetomechanical ratio,” or the extent to which the material
twists in response to changes in its magnetic state. The magnetomechanical ratio is related to another important parameter,
the “g-factor,” a measure of the internal magnetic rotation of the
electrons in a material in a magnetic field.
The magnetomechanical ratio and the g-factor are critical in
understanding magnetization dynamics and designing magnetic
4`J.X. Przybysz, S.P. Benz, C.A. Hamilton, and A. Worsham. U.S. Patent
5,812,078 (issued 9/22/98) “Josephson Junction Digital to Analog
Converter for Accurate AC Waveform Synthesis.”
materials
for
data
storage and spintronics
applications, but they
are extremely difficult
to determine accurately
because of many potential complicating effects.
The NIST experiments
In NIST's Einstein-de Haas experiment, the movements of a canprovide a proof-oftilever were measured with an
concept for using the
optical-fiber laser interferometer.
Einstein-de Haas effect
The optical fiber is 125 micrometers
in diameter, and the end is posito determine the magnetioned less than 10 micrometers
tomechanical ratio and
from the cantilever surface.
the related g-factor in
thin ferromagnetic films. The researchers note that a number of
improvements are possible, such as operating the cantilever
system in a vacuum to reduce the effects of any changes in temperature.
For more information, contact John Moreland, NIST:
[email protected]
NIST Releases New Standard
for Semiconductor Industry
A wide range of optical electronic devices, from laser disk
players to traffic lights, may be improved in the future thanks
to a small piece of semiconductor, about the size of a button,
coated with aluminum, gallium, and arsenic (AlGaAs).
The 1-centimeter square coating, just 3 micrometers thick, is
the first standard for the chemical composition of thin-film
semiconductor alloys issued by the National Institute of Standards and Technology (NIST). Standard Reference Material
(SRM) 2841 was requested by the compound semiconductor
industry to help measure and control thin film composition as
a basis for optimizing material and device properties. The SRM
can be used to calibrate equipment for making or analyzing
these materials. Buyers are expected to include companies that
grow or characterize thin films or use them to make devices, as
well as government and university laboratories.
AlGaAs is used as a barrier material to increase conductivity
in high-speed circuits for wireless communication; semiconductor lasers for optical disk drives, bar code scanning, xerography,
and laser surgery; and light-emitting diodes for remote controls,
traffic lights, and medical instruments. The NIST standard is
expected to increase the accuracy of chemical characterization
of AlGaAs films by an order of magnitude over the current state
of the art. Improved accuracy will reduce wasteful duplication
of reference wafers, increase the free exchange of thin-film
materials between vendors and their customers, and ultimately
improve the accuracy of data on relationships between material
composition and properties.
SRM 2841 can be ordered at http://ts.nist.gov/ts/htdocs/
230/232/232.htm
5 T.M. Wallis, J. Moreland and P. Kabos. “Einstein-de Haas effect in a
NiFe film deposited on a microcantilever,” Applied Physics Letters,
September 18, 2006.
8
|
MEASURE
More NMI News on page 10
www.ncsli.org
RUNNING HEAD GOES HERE
MEASURE
|
9
NMI NEWS
NIST Time and Frequency Metrology Group
Patents Low-Noise Microwave Oscillator
Scientists at the National Institute of Standards and Technology
(NIST) in Boulder, CO have patented an oscillator
that produces very low noise profiles at microwave
frequencies.
Oscillators are devices that ideally produce a
signal at a very precise frequency which can be
used in electronic devices to tell time or to regulate
a multitude of functions. As technology improves,
there is a continuing need to find precise, costeffective oscillators that work at higher and higher
frequencies. Historically, the most precise oscillators have been complicated and expensive to build
because of the low-power signals that must be used
relative to background noise and the shields and special mounts
that eliminate sensitivity to the environment.
The best microwave oscillators previously available used solid
dielectrics, often coupled with cryogenic systems to reduce
background noise, but could not operate without distortion at
high power levels. To meet the needs of industry, researchers at
NIST took a radically different approach when designing this
new microwave oscillator. The resonator uses an open
chamber, or cavity, that is about the size of a roll of quarters,
rigidly defined by an ultra-stiff ceramic housing. The dimensions are such that any frequency that does not agree with a calDC to 50 GHZ 4/18/06 4:22 PM Page 1
culated,
selected resonant frequency of the cavity is rejected.
High power in an air or evacuated ceramic cavity makes possible a less complex oscillator with less stringent amplifiers, one
of the main sources of oscillator noise. At room temperature,
the NIST oscillator produces the same, if not better, results than
prior laboratory oscillators. The
cavity is small and can be produced using any of several relatively inexpensive, ultra-stiff
materials. The result is a simpler,
low cost, stable oscillator with
reference-standard quality and
high resistance to the environment.
Extremely stable oscillators
have several applications outside
of the research environment. A
low noise, high-frequency signal is essential for high resolution
of radar systems, or for increasing the amount of data sent
through communications satellites. In metrology, these oscillators can be used in atomic spectroscopy and as a reference frequency source for microwave tests and calibrations.
NIST researchers continue to improve the design, focusing on
making it smaller and even more resistant to vibration and
severe environmental conditions. For now, the new design represents a significant advance in the development of oscillator
technology.
Contact: David Howe, NIST, [email protected]
Metrology Services
www.dynamictechnology.com • 810.225.4601
Chicago
10
|
MEASURE
Cleveland
Dallas/Ft. Worth
Detroit
www.ncsli.org
MEASURE
|
11
METROLOGY NEWS
METROLOGY NEWS
New "Springer Handbook
of Materials Measurement
Methods"
This handbook, published by Springer, presents, for the first time, advanced methods for
materials characterization and measurement. Key topics
covered in the 22 chapters, include: Measurement Principles
and Structures; Measurement Strategy and Quality; Methods for
Composition and Structure; Materials Property Methods
(Mechanical, Thermal, Electrical, Magnetic, and Optical); Material Performance (Corrosion, Friction & Wear, Environmental
Interactions, Condition Monitoring, etc.); International Standards; and Modeling and Simulation Methods. The handbook
features over 500 color illustrations, 50 comprehensive tables,
and up-to-date approved references. Included with the handbook is a fully searchable CD-ROM for quick data access and
links to helpful sources. The 1200 page handbook was published in August 2006.
For more information, visit www.Springer.com
IAS Requires CCT for Calibration Lab
Accreditation by 2009
International Accreditation Service, Inc. (IAS) has stipulated
that all IAS-accredited calibration laboratories must have their
technicians certified under the American Society for Quality
(ASQ) Certified Calibration Technician (CCT) program by
December 2009, a move that is expected to raise the bar and
provide great benefit to the industry. IAS believes that since the
requirement is placed on calibration laboratories, then IAS
should lead by example. Hershal Brewer, who leads the IAS
accreditation program for calibration, was recently recertified
by ASQ under the CCT program.
For more information, visit www.iasonline.org
CCT Success Story in Tennessee
Arnold Engineering Development Center
(AEDC), located on the Arnold Air Force Base in
Tennessee, now has 15 of their group of 21 specialists earning the American Society of Quality (ASQ)
certified calibration technicians (CCT) credentials. That represents over 70% of their calibration personnel. After the first
specialists received their CCT certification, they helped teach
the most recent preparatory class of six. “All of these individuals either studied and took the course on their own time or went
through Motlow State Community College on their personal
time,” said David Compton, the manager of the Precision Measurement Equipment Laboratory (PMEL). “They took the initiative to do this, and I salute each of them for reaching this
milestone." Jerry Erickson, part AEDC’s PMEL who recently
earned his CCT credentials, said: "To qualify for this certifica12
|
MEASURE
tion and recognition, one must put in some serious study over
a period of several months in subjects such as general metrology, measurement systems, calibration systems, applied mathematics, quality systems and uncertainty. The national pass
average is about 60 percent, so we are proud to state that the
PMEL is at a 100 percent pass rate." All of the center’s newest
CCT-qualified individuals are employees of Aerospace Testing
Alliance (ATA), the support contractor for AEDC. ATA is a
joint venture of Jacobs Sverdrup, Computer Sciences Corp. and
General Physics Corp.
For more information about AEDC: www.arnold.af.mil
LXI Consortium Applauds Web Log Hosted
by Test & Measurement World Magazine
The LXI Consortium is alerting test-system developers to the
availability of the “LXI: Instruments and Applications” Web log
(blog) hosted by Test & Measurement World magazine. Moderated by Chief Editor Rick Nelson, the blog can be found at:
www.reed-electronics.com/tmworld/blog/1330000133.html
The blog’s first entry
was posted in April 2006.
One key purpose of the
blog is to keep readers up
to date on the LAN eXtensions for Instrumentation
(LXI) standard. LXI is the
LAN-based successor to
GPIB. The LXI standard
goes beyond GPIB to
provide additional capabilities that reduce the time it takes to set up, configure and debug
test systems. LXI also helps integrators leverage the time and
effort already invested in system software and architecture. The
standard is managed by the LXI Consortium, a not-for-profit
corporation comprised of leading test and measurement companies. The group’s goals are to develop, support and promote the
LXI standard. LXI’s flexible packaging, high-speed I/O, and
prolific use of LAN address a broad range of commercial, industrial, aerospace and military applications.
The site also provides a forum for visitors to ask questions or
share their expertise in LXI instrumentation. Recent postings
highlight several LXI-related resources, including a series of
application notes and an online LXI tutorial created by UKbased www.radio-electronics.com.
“We at Test & Measurement World believe it’s important to
keep our readers informed about emerging standards and technologies, their benefits and limitations, and the products that
conform to them,” said Nelson. “With the emergence of LXI, we
thought the blog format would be a great way to post timely information and to encourage readers to share their LXI experiences.”
“The LXI Consortium applauds Rick Nelson and the magazine for creating an online forum focused on the LXI standard,”
said Bob Stasonis, co-chair of the Consortium’s marketing committee. “We encourage everyone interested in LXI to visit the
site and contribute to the online discussions.”
Continued on p. 14
www.ncsli.org
© 2006 Northrop Grumman Corporation
9mlgeYl]\ [YdaZjYlagf kg^loYj]
oal` Y hmdk]&
You probably know SureCAL® for its leading calibration software — products that help you
maintain your instruments to exacting precision. But we also have a human side, whose hallmark
is unparalleled customer service. We believe in the value of an intelligent voice, so we feature
live customer support. It’s one more way SureCAL® provides high technology with a human touch.
www.northropgrumman.com Search: SureCAL
METROLOGY NEWS
Additional information about LXI-compliant products, as
well as licensing, specifications and consortium membership, is
available at: www.lxistandard.org.
For more information contact Bob Rennard, President of LXI
Consortium, [email protected]
NACLA Announces Operational Changes,
Restores Recognition of Two Accreditors
The National Cooperation for Laboratory Accreditation
(NACLA) has made several recent changes to its operations in
order to provide better services to U.S. specifiers, regulators,
laboratories and accreditation bodies.
With the approval of the current signatories of the NACLA Mutual Recognition
Arrangement (MRA), NACLA terminated
the MRA effective August 4, 2006. This
change was made so that NACLA could
offer expanded recognition programs, tailored to meet the needs
of government and industry. The NACLA evaluation process will
now focus more on customers, but will continue to be based on
international and national standards.
Following the elimination of the MRA, NACLA made two
additional announcements on September 21, 2006. First, it has
reinstated two major accreditation bodies (ABs) to the list of
NACLA-recognized ABs, and second, it has revised the composition of its Acceptance Panel.
14
|
MEASURE
The reinstated ABs are IAS (International Accreditation
Service) and NVLAP (National Voluntary Laboratory Accreditation Program). Both had voluntarily withdrawn from the
NACLA MRA, which necessitated their removal from the
NACLA recognition roster. As in the past, NACLA continues to
grant recognition solely on the basis of an AB’s demonstrated
competence and compliance with both the NACLA requirements and the accepted international standards (ISO/IEC
17011 and 17025). The competence of IAS and NVLAP was
amply demonstrated through NACLA evaluations, so their
recognition has been restored. With these reinstatements, there
are now seven ABs that are recognized by NACLA.
“It is important for all NACLA stakeholders to understand
that the elimination of the NACLA MRA in no way diminishes
the technical rigor or international validity of NACLA’s evaluation process,” according to Dr. William J. Tilstone, NACLA’s
President. “The standard of competence that has been used in
NACLA’s evaluation program from the start will still be applied.
The main difference from our previous MRA-driven system is
the lack of a requirement that the AB, once its competence has
been ascertained, will no longer be asked to sign a NACLA
MRA.”
The second change occasioned by the elimination of the
NACLA MRA is a restructuring of the Acceptance Panel. This
was done to increase the proportion of government and industry representatives on the Panel, thus giving users of accredited
laboratory results more input into the recognition process. “At
www.ncsli.org
METROLOGY NEWS
the end of the day, the clients for laboratory data are government agencies and private-sector companies,” Dr. Tilstone said.
“It is appropriate, therefore, that representatives of these organizations should have the primary role in deciding which ABs
merit recognition.”
For more information: www.nacla.net
Agilent Labs’ Len Cutler
Leaves a Lasting Legacy
Agilent Technologies mourns the loss of Len Cutler, Agilent Distinguished Fellow and member of the technical staff in Precision
Instrumentation at Agilent Laboratories. After a career of 57
years, Len passed away September 4, 2006 while camping with
his family in Big Basin, California. He was an internationally
recognized inventor who made significant contributions to the
worlds of science and technology, particularly in precision time
measurement and laser interferometry.
In 2004, Agilent Laboratories promoted Len to Agilent Distinguished Fellow in recognition of his long-standing and farreaching contributions that have had, and will have, an
enduring impact on the company. Len was the first and only
person to hold this position, Agilent’s highest technical honor.
“We honor Len for leaving a lasting legacy, and acknowledge
his leadership as an innovator and researcher at HewlettPackard and Agilent for almost 50 years,” said Darlene
Solomon, chief technology officer and vice president Agilent
Laboratories. “Len set the standard for world class research; he
served as a mentor to so many of our engineers and scientists at
Agilent Labs, and remarkably, continued as an active contributor to our research program until his passing. We have lost a
great man, a brilliant researcher, a wise leader and a good friend.”
Len has been aptly named “Father Time.” Over the past 40
years, his innovations and inventions have led to the world’s
most accurate commercial time keeping devices, beginning with
the first solid-state atomic clock in 1964, and leading up to the
Hewlett-Packard 5071A cesium clock, introduced in 1992, with
an accuracy of one second in every 1.6 million years. Clocks
designed by Len form the cornerstone of the time standard
maintained by laboratories throughout the world. His clocks
were the first to be flown in airplanes to perform the synchronization of world clocks and later to establish the variations in
the flow of time predicted by Albert Einstein. The impact of this
work is crucial to modern commerce. Without accurate time
keeping, there would be no GPS navigation, modern computer
networks would no longer function, and financial transactions
would grind to a halt.
Lyons and Sherwood built the first atomic clock in the 1950’s
at the National Bureau of Standards based on theoretical work
by Maxwell and Rabi. Pioneering work by Townes, Zacharias,
Essen, and Ramsey then led to the first cesium beam clock. Len
Cutler began his work on atomic clocks at Hewlett-Packard in
1959, introducing the first solid state cesium beam clock in
Continued on page 17
MEASURE
|
15
1964. The performance and reliability of these clocks were
much better than anything previously available.
Over the years, Len and his team made many improvements
including fundamentally new techniques, like trapped ion frequency standards, and optically pumped atomic clocks, as well
as many contributions to the optimization and theoretical
understanding of the cesium clock.
He has authored a total of 25 patents in many areas of science
and technology. Perhaps his most important invention is his
method for the precise measurement of distance using a two-frequency laser interferometer system. This invention is the crucial
element in the step-and-repeat lithography systems used for the
manufacturing of silicon integrated circuits, in which nanometer resolution is required. Len’s reputation for innovation
resulted in his consultation on the high-visibility failure of one
of the nation’s premier high-technology rapid transit systems in
the early 1970’s. After a serious accident on a major U.S. transit
system, he and his colleagues were quickly able to invent a
patented logic safety system to prevent future incidents.
In addition to his scientific and technical contributions, Len
was one of the founders of Hewlett-Packard Laboratories and
was the leader and mentor of several generations of researchers
in that lab and its successor, Agilent Laboratories.
NCSLI, NIST, and CPEM Sign Agreement
Reprinted from: www.agilent.com/labs/news/2006features/
fea_cutler.html
For more information: www.icpem.org/
The Conference on Precision Electromagnetic Measurements
(CPEM), NIST and NCSLI have signed a sponsorship agreement under which CPEM2008 will be cohosted by the NCSLI
and NIST Boulder Laboratories. CPEM2008 will be held in
Broomfield, CO, near Boulder, June 8–13, 2008. The proximity of NIST Boulder and the NCSLI Business Office allows the
burden of running CPEM2008 to be easily shared between NIST
Boulder and NCSLI. Basically, NIST will be responsible for all
aspects of the conference excepting the finances, which will be
handled by NCSLI.
This agreement builds on the long historical relationship
between CPEM and NCSLI, both having been founded in
Boulder in the 1950s as a result of metrologists seeing the need
for better communications throughout government and industry. The CPEM holds a conference every two years as a means
of disseminating information concerning precision electrical
measurement principles and techniques relevant to standards
and measurements, the science underlying them, and their
application to practical measurement problems. CPEM2006
was held this summer in Torino, Italy, and CPEM2010 will be
held in Korea.
MEASURE
|
17
We’re in your neighborhood
Let us pick up and deliver — or calibrate your equipment at your site
With six regional labs, it’s easy to access Agilent’s calibration
services. And when the original manufacturer calibrates your instruments, you can be sure that they retain their “like new” performance
Remove all doubt
and accuracy.
Your instruments restored to
like-new condition, returned on time
Thorough, high quality calibration starts with automated, factory-
• Premier measurement expertise
• NIST-traceable metrology
• A2LA-accredited calibration available
written procedures. All necessary adjustments are included, no
matter how many test points are out of cal—and you’ll receive comprehensive pre- and post-adjustment data. We’ll even advise you of
the latest hardware and firmware updates, clean and lubricate parts
that need it, and make minor repairs—all at no extra cost.
For more information, send email
to [email protected] or call
1-800-829-4444 and we’ll connect
you to your local cal center:
Arlington Heights, IL
Bethlehem, PA
Chelmsford, MA
Durham, NC
Irvine, CA
Richardson, TX
www.agilent.com/find/removealldoubt
© Agilent Technologies, Inc. 2006
Instrument availability is more predictable too. Turnaround time from
your local cal center is normally five business days, and if you’re
within a two-hour radius of your local cal center, pick up and delivery
are free. Or when downtime is critical, our Volume Onsite Calibration
(VOSCAL) service brings Agilent’s expertise, equipment and personnel to your site.
To learn more, drop us a line at [email protected]. We’ll reply
with an overview of your nearest cal center’s capabilities and a copy
of our “Remove all doubt” brochure.
2007
NCSL International Workshop & Symposium
Metrology’s
Products
Impact Services
JULY 29 – AUGUST 2
Saint Paul RiverCentre
Saint Paul, Minnesota
on
and
Every product and service that consumers use is
highly dependent on metrology. From the fit and
finish of our vehicles to weights and volumes of
products purchased in the grocery, we are
impacted at every level.
www.ncsli.org/conference
[email protected] • [email protected]
303-440-3339
[email protected]
Metrology laboratories calibrate equipment used
to create compatible component parts used in
commercial and consumer products. A sound and
cohesive metrology and quality system, from the
National Metrology Institute to the end consumer,
impacts the quality of life for everyone.
20
|
MEASURE
www.ncsli.org
QUAMETEC INSTITUTE OF
MEASUREMENT TECHNOLOGY
METROLOGY & ISO17025
CONSULTING SERVICES
COURSES IN:
Metrology & ISO17025
• Public Classes
• Self-Paced CD/DVDs
• Onsite Training
• Coming Soon:
eLearning Courses
•
•
•
•
•
•
Quality Systems
Gap Analysis
Compliance Audit
Uncertainty Analysis
Metrology Consulting
Instrumentation
QUAMETEC PROFICIENCY
TESTING SERVICES
MEASUREMENT
QUALITY VALIDATION
•
•
•
•
•
Small Uncertainties
Quick Test Results
PT Planning
Custom Test Design
A2LA Accredited
CALL FOR MORE
INFORMATION
CALL FOR MORE
INFORMATION
CALL FOR MORE
INFORMATION
810.225.8588
810.225.8588
260.244.7450
MEASURE
|
21
SPECIAL REPORT
The CIPM Working Group
on Metrology of Materials
S eton B en net t and Grah am Si ms
A b s t r a c t : Following international discussion of the traceability issues arising in the measurement of materials prop-
erties, the Comité International des Poids et Mesures (CIPM) in October 2005 accepted the proposal to set up an ad
hoc Working Group on Metrology of Materials (WGMM). The WGMM is assessing a wide range of materials properties, looking particularly at the need for improved traceability routes, data comparability, and the availability of
appropriate reference materials. The Working Group on Materials Metrology will report to CIPM in October 2007,
with the intention of raising the profile of materials metrology internationally and engaging the leading National Measurement Institutes in recognising and addressing known difficulties in demonstrating traceability of many material
properties to the Système International of units (SI). Terms of Reference have been agreed, and the first meeting took
place at the United Kingdom’s National Physical Laboratory in May 2006. This paper describes the range of properties being investigated and highlights the studies being undertaken by the WGMM members.
1. M e a s u re m e n t s , Tr a c e a b i l i t y
a n d S t a n d a rd s
In most areas of metrology, the concept
of traceability to established national or
international standards is well understood. The SI (Système International)
provides a coherent set of well-defined
units which provide a common language
for expressing and understanding the
results of measurements. The national
metrology institutes (NMIs) maintain
Seton Bennett and Graham Sims
National Physical Laboratory
Teddington TW11 0LW
United Kingdom
22
|
MEASURE
standards according to their national
needs and the equivalence of these standards has been established through the
mechanisms of the Comité International
des Poids et Mesures (CIPM) Mutual
Recognition Arrangement which includes:
Key intercomparisons, mutual review of
claimed capabilities, and regional assessment of NMI quality systems.
The result of this methodology is that
NMIs can generally demonstrate excellent agreement in intercomparison exercises in these classical fields of
measurement, and they pass this metrological confidence on to accredited
testing and calibration laboratories
which are in turn required to demonstrate the traceability of their results.
Figure 1 shows the results of measure-
ments of the length of a 175-millimetre
gauge block by eleven NMIs. [1] All the
results agree within 1 part in 106 and,
with one exception, the agreement is
better than 3 parts in 107. This impressive result reflects a very thorough and
careful approach to the relatively
straightforward measurement of a length
standard.
Similar exercises to evaluate the ability
of laboratories to measure materials
properties are often less impressive.
Figure 2 illustrates the results of a
national intercomparison exercise in the
UK to determine Young’s modulus. [2]
The results from 25 laboratories for a
simple modulus measurement show a
spread of nearly 25 %. There are many
possible reasons for this enormous diswww.ncsli.org
SPECIAL REPORT
150
100
50
0
–50
-100
–150
–200
CNR-IMGCT
PTB
NPL
NIST
INMETRO
NRC
NMIJ
NIM
CSIRO-NIML CSIR-NIML
VNIIM
F i g u re 1. Results of Consultative Committee for Length (CCL) Key Comparison of 175-mm
gauge block. [1]
25
12.5
0
–12.5
N04
N13
N07
N16
N05
N06
N01
N22
N08
N10
N11
N24
N02
N18
N12
N17
N09
N19
N03
N20
N15
N25
N14
–25
N26
The need for wider international collaboration in the measurement of materials
properties, and in particular the issues of
standards and traceability, have been
under discussion for a number of years at
meetings of the Versailles Project on
Advanced Materials and Standards
(VAMAS). [3] VAMAS operates under a
Memorandum of Understanding signed
by senior representatives of government
in countries of the Economic Summit
(G7) and of the European Community. It
supports international trade through
projects aimed at providing the technical
basis for drafting codes of practice and
specifications for advanced materials.
The scope of this international collaboration embraces all aspects of enabling
science and technology required as a precursor to the drafting of standards.
Through its activity, VAMAS fosters the
development and harmonisation of international standards for advanced materials by the various existing standards
agencies. Since its inauguration in 1982,
VAMAS has had a considerable impact
on the development of internationally
1 7 5 mm S te e l Ga u ge Bloc k , S /N 6 0 7 1
200
N23
2. I n t e r n a t i o n a l C o n c e r n a nd
th e R ol e o f the C IPM
CC L - K 2
D e g re e s o f e q u i v a l e n c e [ D j 1 a n d i t s e x p a n d e d u n c e r t a i n t y ( c o v e r a g e f a c t o r : 2 ) U j i ]
Va r i a t i o n i n You ng ' s M o d ul us , %
crepancy between laboratories, but the
contrast with the level of agreement
obtained by NMIs comparing results for
the calibration of a length standard
could hardly be more marked.
A key question is to differentiate
between intrinsic properties of materials
and other parameters related to the
form and scale of a material specimen.
Thus thermal expansion or Young’s
modulus are quite clearly intrinsic properties, while surface finish or particle size
describe individual samples and may
effect the results of measurements of
properties. The use of standardised
measurement procedures, as for hardness, creates a repeatability which
depends on careful adherence to the
accepted measurement sequence and a
form of traceability when every laboratory uses the same procedure. Separating
the properties of materials from the
factors and problems which influence the
results of measurements is a clear prerequisite for any study of genuine material property measurement traceability
issues.
L ab o ra t o ry Co d e N u mbe r
F i g u re 2. Results of round robin determination of Young’s modulus. [2]
accepted standards for engineering materials. The specification of materials in
terms of their characterisation and their
performance is based mainly on measurement methods and procedures, with a
lack of emphasis on the need for reliable
traceability.
In 2004, the VAMAS Steering Committee approved actions aimed at bringing about closer collaboration between
VAMAS and CIPM. Seeing the need to
widen participation in their activities, Dr.
Graham Sims (VAMAS Chairman) wrote
to Professor Andrew Wallard, BIPM
Director, drawing his attention to the
need to include materials in the formal
international structure for metrology by
engaging the attention of NMIs and the
CIPM. They saw this as the way to bring
international authority and metrological
experience to the hugely important and
growing area of materials metrology.
Following discussion at the meeting of
NMI Directors in September 2004,
MEASURE
|
23
SPECIAL REPORT
BIPM hosted a workshop in February 2005 to explore the
issues, identify specific traceability questions in materials
science and propose further international initiatives in the field.
The conclusions of this workshop informed a discussion at the
2005 meeting of the CIPM, which decided to set up an ad hoc
Working Group on Materials Metrology (WGMM).
P ro p e rt y A re a
Detailed Aspects
Mechanical Properties
Hardness
Modulus
Strength
3. T he C I P M Wo r k i n g G ro u p o n M a t e r i a l s M e t ro l o g y
Following the 2005 CIPM decision, the ad hoc Working Group
on Materials Metrology has been established with experts from
NMIs and other institutes in some 10 countries. Its Terms of
Reference are to:
1. Identify those material properties for which globally comparable, traceable measurement results are important for
science, engineering and manufacturing technology;
2. Identify those material properties for which the needs for
traceable measurements are not covered by the activities of
the Consultative Committees;
3. Establish the user needs for activity in materials metrology;
4. Investigate the existing capabilities of participating NMIs
by initiating some pilot studies, including a small number
of interlaboratory comparisons;
5. Develop tools and methodologies for establishing traceability in materials testing;
6. Define the objectives, aims and initial activities for an
ongoing programme in metrology for materials, including
recommendations for underpinning activities, such as the
organisation of Key Comparisons and the development of
Reference Materials and Reference Methods;
7. Liaise closely with other interested organisations; and
8. Report its conclusions to the CIPM by October 2007.
Toughness
Fatigue
Creep
Viscosity
Thermophysical
(Phys-Chem)
Properties
Conductivity
Diffusivity
Expansion
Specific Heat
Emissivity
Composition and MicroStructural Properties
Grain size / boundaries
Phase
Porosity
Texture
Particle size
Defects
Functional Properties
Electrical
Optical
Magnetic
Thermo-electric
4 . In iti al S tep s
The first meeting of the Group, in May 2006, began with a discussion of the materials properties which are important for
science and manufacturing, and explored some of the issues
associated with establishing traceability to appropriate standards. To the list of obvious properties (mechanical, electrical
and thermal coefficients of solids) were quickly added a number
of properties of liquids and the distinct properties of materials
on the nano-scale (see Table 1).
No experimental studies or intercomparisons are planned at
this stage, as the Working Group on Materials Metrology
decided it was more important to concentrate on collecting
information about previous exercises and to study existing provisions for traceability when materials properties are being
measured in testing laboratories and elsewhere. The Group
recognised that in some cases traceability may be to a standard
or a procedure, rather than to the SI in the generally accepted
sense, and that the reliability, repeatability and reproducibility
of results will be affected by a number of factors. There is also
a need to seek the views of the user community in order to identify those properties for which repeatability and comparability
are particular problems.
Table 2 identifies some of the user aspects and needs for comparability of data that are used in support of regulations and
certification.
24
|
MEASURE
Acoustic
Electrochemical
Properties
Ta bl e 1. Material properties to be initially assessed.
5. The Wa y A h e a d
The Working Group on Materials Metrology will meet again in
December 2006. Meanwhile, as well as collecting historical data
about previous intercomparisons and identifying issues associated with specification standards and measurement procedures,
an attempt will be made to obtain views and information from
the user community, including answers to the following questions:
• For which material properties is it particularly difficult to
demonstrate traceability as the “property of the result of a
measurement . . . whereby it can be related to stated references, usually national or international standards, through
an unbroken chain of comparisons all having stated uncertainties?” (definition from the International Vocabulary of
Metrology (VIM), 1993)
• For which measurement results is it difficult or impossible to
express in terms of SI units?
• Where is there a particular need for intercomparisons to
www.ncsli.org
SPECIAL REPORT
A sp e c t
For each property (e.g., Modulus)
Scope
Short description of property and measurement
methods/techniques.
Material Category
Meta, ceramic, polymer, composite, rubber, etc.
Material State and Scale
Liquid, micro, particle, bulk, film, nano, surface
Regulation Need
Is this property a directly or implied requirement in
any regulative document (e.g., law; EU directive;
industry (CAA, DNV, Lloyds) regulation; etc.)?
User Needs
What are the user requirements (excluding the
regulatory aspects listed above)?
Accreditation Needs
Is this property needed in any accreditation
procedure?
Comparability
Is there a need for intercomparability of data (e.g., if
required in design codes, when more than one code
and/or test method exist for the same application,
such as pressure vessels)?
Need for
Intercomparison
Should the WGMM initiate intercomparisons or
recommend future exercises?
Economic Impact
Studies
Have any economic studies been undertaken of the
impact of incorrect or high scatter measurements of
this property?
Number of
Existing Methods
How many “standardized” or accepted methods are
used to measure this property?
Standardisation
Situation
What standards exist and do they include precision
data?
NMI or Metrology
Research Institute
Activities
Is there any activity in this property measurement at
these organizations?
Research &
Development Phase?
Is there any basic research into this property and/or
new measurement techniques?
Traceability Issues
What is the traceability route for this property?
SI units Relevant
Which SI unit (if any) is relevant?
CCs Coverage
Is this property covered by any existing BIPM
Consultative Committee (CC)?
Prior Studies
(Method Specific)
Have any prior studies been undertaken for this
property?
Method Comparability
Have any prior comparability studies been
undertaken for this property?
Standard Test Machines
Do standard test machines exist for these
measurements?
establish equivalence and repeatability
between different laboratories and/or
various procedures?
Following the December 2006 meeting,
the Working Group on Materials Metrology will undertake further work, possibly
including some very limited pilot intercomparisons, before preparing its report
for the CIPM meeting in November
2007. This report may include recommendations for new initiatives to
improve the comparability and traceability of the measurement of materials properties worldwide.
6. Conclus ion
This paper is based on a presentation to
the NCSLI Workshop & Symposium in
Nashville in 2006. Our wish is to put
these questions to a wider audience, and
we invite readers in testing laboratories,
accreditation bodies, materials producers
or manufacturers, who have a need to
obtain accurate values for the properties
of the materials they use, to contact the
Working Group on Materials Metrology.
Information about specific requirements
and case studies on past difficulties are
of particular value, especially when supported by references or detailed measurement data. You can contact us by e-mail at
[email protected] or
[email protected].
7 . Re f e re n c e s
[1] A. Lewis, CCL-K2, “Long Gauge Block
Measurement by Interferometry: Final
Report,” Metrologia, vol. 40, Tech.
Suppl., no. 04004, 2003.
[2] B. Roebuck, J.D. Lord, P.M. Cooper and
L.N. McCartney, “Data Acquisition and
Analysis of Tensile Properties for Metal
Matrix Composites,” ASTM J. Testing
and Evaluation, JTEVA, vol. 22(1), pp.
63-69, 1994; and presented at the ASTM
Workshop on Accuracy of Load and
Strain Measurements, Miami, FL, Nov.
18, 1992.
[3] For more information on VAMAS, see
the web site www.vamas.org.
Have machines been compared?
Calibration – Reference
Materials
Do calibration reference materials exist for this
property?
Tabl e 2. Aspects of material property use, traceability and support to regulations.
MEASURE
|
25
SPECIAL REPORT
Requirements for the Calibration of
Measuring and Test Equipment
D el Caldwell
A b s t r a c t : This paper provides an introduction to the new ANSI/NCSL Z540.3 standard and its approach to prescrib-
ing requirements for a calibration system that controls the accuracy of the measuring and test equipment used to ensure
that products and services comply with prescribed requirements. The new ANSI/NCSLI Z540.3 standard replaces Part
II of the current standard, ANSI/NCSL Z540-1 (R2002). Z540.3 consists of six clauses, each of which is described:
Scope; References; Terms and Definitions; General Requirements; Calibration System Implementation; and Calibration System Assessment and Improvement. This paper also discusses the three requirements that received the most
interest and discussion during the development process: (1) Measurement decision risk criteria; (2) Test uncertainty
ratios; and (3) Use of calibration laboratories accredited to ANS/ISO/IEC 17025.
1. I nt ro d u c t i o n
On 3 August 2006, NCSL International received a notice from
the American National Standards Institute (ANSI) that
ANSI/NCSL Z540.3-2006, Requirements for the Calibration of
Measuring and Test Equipment had been approved by ANSI’s
Del Caldwell
Caldwell Consulting Group
906 Pomona Ct.
Claremont, CA 91711-3864 USA
26
|
MEASURE
Board of Standards Review as an American National Standard.
This announcement culminates the efforts of Working Group
One of NCSLI’s 174 Standards Writing Committee that began
in early 2003.
ANSI/NCSL Z540.3 “prescribes requirements for a calibration system to control the accuracy of the measuring and test
equipment used to ensure that products and services comply
with prescribed requirements.”
The new American National Standard is intended to replace
Part II of the current standard, ANSI/NCSL Z540-11994(R2002) [1], which addressed compliance requirements
for a calibration system for measuring and test equipment. With
the recent adoption by ANSI of ISO/IEC 17025, “General
www.ncsli.org
SPECIAL REPORT
Requirements for the Competence of Testing and Calibration
Laboratories” as an American National Standard [2], the basic
functionality of Part I of the current standard, is also replaced.
Part I of the current standard includes a mix of compliance and
competency demonstration requirements for a calibration or standards laboratory. As a result of these two actions, the stage has
been set to rescind the current Z540-1 standard in mid-2007.
2 . Pu r p o se o f Z5 4 0 . 3
In many organizations, measuring and test equipment are used
in the research, development, test, evaluation, production, and
support of products and services to a wide range of customers.
The information gained from the use of the measuring and test
equipment contributes to our knowledge of the organization’s
product or service and to the associated decisions about their
quality and suitability for their intended application. The validity of measurement results are significantly affected by the
uncertainty of the measuring and test equipment, which is of
particular importance to the success of the organization.
This new American National Standard describes a management system used to ensure the continued accuracy of measuring and test equipment used to support the organization’s
endeavors. This approach is described in the form of a calibration system or program. Measuring and test equipment that
affect the quality of the organization’s product or service and
their conformity to requirements need to be included in the calibration system. Performance requirements for the measuring
and test equipment are determined and subsequently used as a
basis or guiding criteria for the system’s calibration support of
the equipment. The measuring and test equipment are periodically calibrated using a defined process that ensures that the calibration results are traceable to the International System of
Units (SI), usually through the U.S. National Institute of Standards and Technology. Monitoring and controlling the performance of the equipment are part of the calibration system’s
functions. As a result, personnel using the calibrated measuring
and test equipment should be able to place confidence in the
performance of the equipment included in the system.
The new Z540.3 is a requirements oriented document and
suitable for use in contract applications. Therefore “how to”
type information is limited to guidance in notes, as implementation of the new Standard may vary within and among organizations. However, additional descriptive information is planned
to be included in an NCSLI Handbook that is currently under
preparation by the NCSLI 171 Calibration System Resources
Committee.
3 . C on t en t s o f Z5 4 0 . 3
ANSI/NCSL Z540.3 is comprised of six major clauses:
Clause 1: Scope
Clause 2: References
Clause 3: Terms and definitions
Clause 4: General requirements
Clause 5: Calibration system implementation
Clause 6: Calibration system assessment and improvement
The basic content of each clause is described in the following
paragraphs. Figure 1 is provided to illustrate the implementaVol. 1 No. 4 • December 2006
tion and effect of Clauses 4, 5, and 6.
The scope is included in Clause 1 and has been described in
the previous introductory statements above.
There are two types of references included in Clause 2 of the
new Standard: Normative and Informative. One normative reference is included, the VIM, to provide a reference for terms not
defined in the Standard. Three informative references are
included that apply only to the extent described. These include:
ANSI/NCSL Z540-2-1997 (R2002) (the equivalent of the ISO
GUM) [3]; ANS/ISO/IEC 17025:2005 [2]; and NCSLI RP-11996 [4]. These documents provide supplemental information
on the expression of measurement uncertainty, calibration laboratory competency assessment, and calibration interval establishment and adjustment, respectively.
There are thirteen terms and definitions that have been
included in Clause 3. These terms and associated definitions
were included where either the definition in the VIM needed to
be adapted to reflect common U.S. usage or the term was not
defined in the VIM. Some of the terms are common but can
have multiple meanings in everyday usage. Accordingly, they
were included to be specific about their application in the new
Standard.
Clause 4 addresses the general requirements that apply
throughout the calibration system and an organization’s implementation of that system. It includes the fundamental objective:
“establish, document, operate, and improve a system to
manage the calibration of measuring and test equipment. The
organization shall identify and include measuring and test
equipment in the calibration system having an influence on the
quality of the organization’s product and its conformity to determined requirements.” Of course, the clause also includes the
requirements for identifying the performance requirements of
the measuring and test equipment based on their application
and using this information as the principal drivers for the technical aspects of the calibration system, such as, measurement
reliability, calibration service quality, calibration tolerances, etc.
This clause also includes requirements for quality objectives,
personnel, information resources, and shipping and handling of
measuring and test equipment.
Clause 5 describes the technical requirements for implementing and sustaining a calibration system. This clause includes the
following components:
5.1 Calibration requirements. This sub-clause basically reiterates the requirement for organizations that utilize measuring
and test equipment to provide a product or service are required
to identify those equipments that have an influence on the
quality of the organization’s product or service and include them
in the calibration system.
5.2 Measuring and test equipment. The focus of this subclause is the requirements for identification of measuring and
test equipment that are included in the system, including their
calibration status. Management of adjustment means and nonconforming measuring and test equipment are also addressed.
5.3 Calibration of measuring and test equipment. The principal technical requirements for calibration of measuring and test
equipment are addressed in this sub-clause. Included are
requirements for:
MEASURE
|
27
SPECIAL REPORT
Calibration System
Product-Service:
• Research
Calibration System:
General Requirements
• Calibration System
Implementation
• Calibration System
Assessment and
Improvement
• Development
•
NIST & SI
Calibration Quality
Requirements
• Production
• Test
Calibrated
M&TE
& Inspection
Customer
• Support
• Operations
• Maintenance
Measurement
Reliability &
Calibration Intervals
M&TE
Application
Requirements
Traceability
Suppliers
M&TE Calibration Requirements
Assessment &
Improvement
Calibration Service Requirements
F i g u re 1. Calibration system of an organization controlling measuring and test equipment (M&TE).
•
•
•
•
•
•
•
Measurement uncertainty for reporting measured values;
False accept risk or test uncertainty ratio for tolerance tests;
Calibration procedure objectives, content, and validation;
Measurement assurance procedure objectives and content;
Measurement uncertainty and traceability;
Calibration equipment availability and application;
Calibration personnel competency, supervision, and authorization;
• Influence factors, condition monitoring, and control;
• Calibration quality monitoring;
• Calibration reporting and reports; and
• Calibration records.
5.4 Calibration intervals. This sub-clause addresses requirements for establishing and adjusting calibration intervals of
measuring and test equipment that are included in the system in
order to assure acceptable measurement uncertainty, traceability, and reliability requirements are achieved. The potential for
conflicts of interest is addressed, along with the requirement for
recall of measuring and test equipment at the end of their calibration interval. Emphasis is also placed on measuring process
control intervals.
5.5 Outside suppliers. Z540.3 addresses requirements for
outside suppliers if their product or service have an effect on the
calibration system. Further, if the outside supplier is acting as a
8
component
of or supplements the calibration system, the supplier is required to meet the applicable requirements of the new
Standard. For example, if the organization contracts with an
outside supplier to perform measurement uncertainty analysis
28
|
MEASURE
for calibrations being performed within the system, then that
supplier must meet the applicable parts of the new Standard.
Clause 6 addresses the requirements for assessment of all
aspects of the calibration system to ensure conformity of the
system with the new Standard, determine the suitability and
effectiveness of the system, and improve the system as needed.
The overall assessment and improvement activities are organized into five areas:
6.1 Management review. A periodic review to ensure the adequacy, effectiveness, and suitability of the system to achieve
established objectives.
6.2 Calibration system audit. An independent audit of the calibration system to established criteria.
6.3 Calibration system monitoring. Monitoring of critical
system parameters including:
• Performance quality of measuring and test equipment;
• Calibration interval suitability; and
• Calibration service quality and traceability.
6.4 Customer assessment, verification, and feedback.
Addresses involvement by the customer in assessment and verification, including use of customer feedback information.
6.5 Corrective and preventive action. Includes the requirements to identify causes of unacceptable performance and to
eliminate the discrepancies.
4. I m p ro v e m e n ts I n c l ud e d i n Z 5 4 0 .3
During the development of Z540.3, members of Working
Group One took a broad look at what the new Standard offered
www.ncsli.org
SPECIAL REPORT
over continued use of Part II of ANSI/NCSL Z540-1. In general,
the new standard provides mechanisms to improve the quality of
measuring and test equipment performance, its management
and the related calibration services by:
• Incorporating experience with the use of previous and associated standards to improve the criteria for a calibration
system and its components;
• Adding the requirement to identify measuring and test equipment application related performance criteria to assure the
compatibility of calibration intervals and calibration services
to application requirements;
• Improving requirements for calibration procedures including:
addressing the objectives of their use, providing tolerance test
criteria, providing measurement risk management criteria and
guidance, clarifying test uncertainty ratio determination,
increasing topics to be addressed, and providing validation
criteria for their suitability;
• Requiring traceability to the SI units and improving compatibility with the U.S. National Measurement System;
• Adding requirements for procedures and documentation of
measurement uncertainty determination and expression to
improve the application to calibration;
• Providing improved criteria and supplemental guidance for
calibration intervals and the use of measurement assurance
processes for management of measuring and test equipment
performance;
• Extending requirements to control operator accessible calibration adjustments;
• Improving requirements for identifying non-conforming
measuring and test equipment and addressing the impact of
their use;
• Providing requirements for optional use of accredited calibration services; and
• Adding requirements for assessment, quality control monitoring, and system improvement.
Of these, three topics raised the most interest and discussion:
measurement risk criteria; test uncertainty ratios; and use of calibration laboratories accredited to ANS/ISO/IEC 17025.
Sub-clause 5.3 addresses requirements for two types of calibrations: one where measured values are reported to the customer; the other for calibrations involving tolerance tests. For
the first type of calibration, the measured values are reported
with a measurement uncertainty that is required to be acceptable to the customer. The second states that “the probability that
incorrect acceptance decisions (false accept) will result from
calibration tests shall not exceed 2% . . .” This is commonly
called false accept risk. An option to this requirement is to
ensure that the test uncertainty ratio is equal to or greater than
4:1. Here, test uncertainty ratio is simply the ratio of the range
or span of the tolerance to twice the 95% expanded uncertainty
of the measurement process used for calibration (this applies to
two-sided tolerances). Note that both approaches require determination of the measurement uncertainty. These two
approaches can provide comparable calibration quality for
workload with mid- to low-measurement reliability. The reason
for using both the false accept risk and the test uncertainty ratio
approaches is that if you have reasonable information about the
workload to determine false accept risk, you have greater
options for achieving the requirements, such as: adjustment of
calibration intervals for different measurement reliability
levels; use of guard bands for management of test tolerances,
etc. However, if you do not have reasonable information about
the workload, then you could still make use of the test uncertainty ratio requirement. It should be noted that by using the
false accept risk approach, there are many scenarios where you
can achieve an acceptable level of risk and yet the test uncertainty ratio is less than 4:1.
Both sub-clauses 5.3 and 5.3.3.2 require that all calibration
servicing components, such as a calibration or standards laboratory or facilities associated with the calibration of measuring
and test equipment that are included in a calibration system, be
competent. Competency includes either meeting the full
requirements of Z540.3 or being suitably accredited to, or
found to be in conformance with, the requirements of
ANS/ISO/IEC 17025, including meeting the requirements of
the overall sub-clause 5.3.
Sub-clause 5.3 acknowledges that accreditation to
ANS/ISO/IEC 17025 may be an acceptable means to recognize
a calibration service’s competence provided that the scope of the
accreditation is compatible with the specific measuring and test
equipment calibration requirements and the additional requirements of this sub-clause are included in the accreditation
process. This should provide reasonable assurance that the goals
of Z540.3 are met.
Sub-clause 5.3 also allows for the customer to accept alternative, independent assessments of conformance, such as audits,
evaluations, and inspections, which clearly validate the calibration service’s capability, competence, and compliance with
Z540.3 to the customer’s calibration requirements. In either
case, all such assessments are to be made for specific calibration
capabilities and competence which match the parameters
included in the customer’s measuring and test equipment calibration requirements and which are listed in documentation which
attests to their authenticity.
5. Sum m ar y a nd C onc lus i on
During the development of ANSI/NCSL Z540.3, Working
Group One initially invested a significant amount of time and
effort to explore modeling the new Standard in the same form
as the current standard, ANSI/NCSL Z540-1. We included in
the model two international standards, ISO/IEC 17025 and ISO
10012, combining extended versions in a single document to
play the roles of Part I and Part II of Z540-1, respectively. We
found that the resulting document was excessively long and, due
to overlaps and other harmonization issues, difficult to comprehend and use, even though the document in itself was technically acceptable. Accordingly, the final approach described here
was determined to be the best approach for expressing the
requirements for the calibration of measuring and test equipment, as well as fulfilling needs that are unmet by other standards and doing it to take advantage of one of the international
standards (ANS/ISO/IEC 17025).
Working Group One was also conscious of the need to
express the requirements in a straightforward and explicit
MEASURE
|
29
SPECIAL REPORT
manner to facilitate their implementation and interpretation. As
with any change, there are costs and benefits with the use of the
new Standard. We believe that the benefits outweigh the modest
costs. One colleague who is associated with a major aerospace
firm adapted their calibration system to the requirements of
Z540.3, even before it was finished. His view was that the
changeover went smoothly and was relatively easy as their practices, for the most part, addressed the requirements.
Z540.3 has been formatted for publication and will be
included in the NCSLI publications CD provided to all NCSLI
member organizations. Z540.3 will also be available for purchase through the NCSLI Business Office, Boulder, CO.
6. Ac kn o wl e dg em e n t
As the Working Group One chairperson, I wish to thank all the
members for the special contributions that each made throughout the development of Z540.3. The explorations and discussions were open and candid, with each member working to
fulfill the goal of serving the U.S. measurement community.
Members and alternate members of Working Group One,
together with their affiliation, included: Dave Abell, Agilent;
Del Caldwell, CCG; James Clark, Boeing Commercial; Marcel
Dubois, Anritsu; Bill Eyler, Agilent; Doug Faison, NIST/
NVLAP; Chet Franklin, DynCorp; Bob Fritzsche, NSWC
Corona; John Grajera, Lockheed Martin; Dan Harper, Harper
Quality; Jerry Hayes, Hayes Technology; Del Knapp, Tektronix;
Ray Kotowski, NASA; Mark Kramer, USMC; Brian Lee,
30
|
MEASURE
Anritsu; Bill McCullough, CSC; Paul Nelson, Raytheon; and
Derek Porter, Boeing Commercial.
We are also indebted to many others in the community for
their contributions, including: Alane Caldwell, IRES; Howard
Castrup, ISG; Dave Deaver, Fluke; Dennis Jackson, NWSC
Corona; and Larry Nielson, SCE.
7. R e f e re n c e s
[1] “Calibration Laboratories and Measuring and Test Equipment –
General Requirements,” ANSI/NCSL Z540-1-1994 (R2002).
Available from the NCSLI Business Office, Boulder, CO, 80301,
USA.
[2] “General Requirements for the Competence of Testing and Calibration Laboratories,” ANS/ISO/IEC 17025:2005. Available
from the NCSLI Business Office, Boulder, CO, 80301, USA.
[3] “U.S. Guide to the Expression of Uncertainty in
Measurement,”ANSI/NCSL Z540-2-1997 (R2002). Available
from the NCSLI Business Office, Boulder, CO, 80301, USA.
[4] “Establishment & Adjustment of Calibration Intervals,” NCSLI
RP-1. Available from NCSLI Business Office, Boulder, CO, 80301,
USA.
www.ncsli.org
SMART/CMS Enterprise Calibration Management Software
YOUR GOAL, OUR SOLUTION
Now available on
SQL Server 2005
and Oracle 10GR2!
AssetSmart by PMSC
2800 28th Street, Suite 109
Santa Monica, CA 90405 USA
310.450.2566
www.assetsmart.com
[email protected]
Scalable from laptop to
mega-enterprise
Compliance Assurance for
ANSI/NCSL Z540, ISO 9000,
SAE AS9000, ISO 17025 and
FDA 21 CFR Part 11
Automated Tracking of
OOT/SOOT notices
Integration Tools connect to calibrators
and data acquisition systems
User Customizable screens
Integrated Equipment Management
module
Tool Crib and Spares Inventory
module
TECHNICAL PAPERS
Te d D o i ro n
A b s t r a c t : Thermal expansion effects are very important in dimensional metrology. In this paper a measurement model,
and associated equations, are developed for the case of a one-dimensional measurement of a steel test gage using a
measuring machine and master gage. After presenting the uncertainty components for this measurement, several
example measurement situations with different levels of temperature control are calculated and discussed. For each
situation, the magnitude of the different sources of uncertainty are compared in order to rationally allocate resources
to improve the overall measurement uncertainty.
In dimensional measurement the uncertainty is often dominated
by the effects of thermal expansion. [1] This paper discusses
these effects, their sources, and the methods used to determine
the uncertainty components. In an extended example, the
thermal uncertainty components for the measurement of a steel
gage on a one-dimensional universal length measuring machine
(ULMM) is derived for different levels of laboratory temperature control and measurement. By increasing the knowledge of
the temperature of the instrument and gages, the uncertainty of
the measurement is dramatically lowered.
and
G(tg) = [1 + αg(tg – 20)] G20 .
(3)
We also have the scale readout, S(ts). What we would like, of
course, is the actual difference in length between M(tm) and
G(tg). However, the scale reading is not correct because the
scale also changes with temperature.
Suppose the temperature is above 20 ºC. The scale is now
longer, and the distance we measure will seem smaller than it
really is. We thus have to correct the scale reading by enlarging
it in proportion to the thermal expansion. Thus, the actual
length difference between the gage and master is Smeas,
2 . Th e M odel
In order to make a length measurement, we must take the actual
(4)
Smeas = [1+ αs(t s – 20)] S20 .
measured values and calculate what the length would be at
exactly 20 ºC (68 ºF). In the most general case, we will have a
Putting these together, we get:
measuring machine with a scale (S), a master gage (M), and a
test gage (G). To make the corrections we must have the tem[1+ αs(ts – 20)] S2 0 = [1 + αg(tg – 20)] G2(5)
0 – [1 + αm(tm – 20)]
perature and coefficients of thermal expansion (CTEs) of each.
[1+ αs(ts – 20)] S2 0 = [1 + αg(tg – 20)] G2 0 – [1+ αm(tm – 20)] M20 .
The actual measurement equation is:
S(ts) = G(tg) – M(tm) ,
(1)
where M is the length of the master gage, G is the length of the
test gage, and S is the apparent difference in the scale readings
for M and G. Each of these depends on their temperature, tm,
tg, and ts, respectively. If we denote each CTE with α, and the
scale is calibrated to be correct at 20 ºC, we find that:
M(t m) = [1+ α m(t m – 20)] M 20
Ted Doiron
Engineering Metrology Group
National Institute of Standards and Technology
Gaithersburg, MD 20899-8211 USA
32
|
MEASURE
(2)
Now we make a small replacement to make the equation
easier to handle; we use the Greek V to stand for (t – 20). We
then get:
(1+ αsVs) S20 = (1+ αgVg) G20 – (1+ αmVm) M2 0 .
(6)
If we solve for the length of the gage, G20, we get:
G20 =
(1+ α mV m)M20 + (1+ α sVs )S20 .
(1+ α g Vg )
(7)
Now, we can simplify this a bit by noting that the second term
in the denominator is much smaller than 1. We can expand
(1 + αgVg)–1 in a Taylor series, and keep only the first two terms,
–1
(1 + αgVg)
≈ ( 1 – αgVg) .
(8)
www.ncsli.org
TECHNICAL PAPERS
Then we get as our equation:
G20 ≈ (1 + αsVs)(1 – αgVg) S20 + (1 + αmVm)(1 – αgVg) M20 . (9)
Again, since all the αV terms are much smaller than 1, we can
multiply out the right hand side and ignore all of the terms of
order (αV)2 and higher. We are left with the final equation for
the length of the gage:
G20 ≈ (1 + αsVs – αgVg) S20 + (1 + αmVm – αgVg) M20 . (10)
How good is this equation? Let us take an extreme case of a
measurement at 0 ºC on plastic. Plastics have very large CTEs,
some nearly 100 #10-6/ºC. Thus, the term
αV ≈ 100 × 10–6/ºC × 20 ºC × L ≈ 0.002 L .
(11)
The second order terms, being (αV)2 ; 0.000 004 L, are
about 500 times smaller. For most laboratory conditions, the
ratio is much larger, and therefore these second order terms are
negligible.
Thus, to make thermal corrections to a general measurement,
we need a number of quantities that are shown in Table 1.
All of these quantities have some uncertainty, u, associated
with them. The thermal error terms are shown in Table 2.
There are also the uncertainties of the scale reading and the
length of the master gage to include in the overall uncertainty
budget. Since we are focusing on sources associated with
thermal expansion, however, we will not say much about these.
Usually the scale reading uncertainty comes from the certificate
or the manufacturer’s specification if it is certified to its specification, rather than being calibrated. If the correction from the
master gage calibration certificate is used, the scale reading
uncertainty is taken from the calibration laboratory’s stated
uncertainty. If the calibration only certifies the gage to an accuracy class or grade, the width of the class or grade tolerance is
taken as a rectangular distribution. [2]
The rest of the quantities are more difficult to estimate; you
actually have to think a bit. In some cases the CTEs can be
determined to some precision from calibration. At NIST we
have calibrations of the CTEs of our gages, either in house or
αs
CTE of the scale of the measurement instrument
αm
CTE of the master gage
αg
CTE of the test gage
ts
from calibration reports from the manufacturer. In these cases
the uncertainty in the CTE can be quite small, 0.1 #10-6/ºC or
better. Without such detailed information you must use whatever information you can get. The range of CTE for “steel” is
quite large, but there are only a few gage block steels, and the
range of CTE for these is somewhat smaller. The Standards for
gage blocks [3] usually require that the CTE of a steel gage
block is 11.5 #10-6/ºC with a tolerance of ±1 #10-6/ºC. Other
materials, such as tungsten carbide, chrome carbide, ceramic,
etc., have no specification and most people will assume ±10 %
as the uncertainty. [4]
The temperatures are more complicated, still. Some notes:
1. If you only have one thermometer, and it is used only to
monitor the room, then you must use the daily variation in
the room temperature as the uncertainty for everything.
This is a very large number, but uncertainty is a measure of
your ignorance of the measurement and if you don’t
measure something you are pretty ignorant.
2. If your thermometer is calibrated you still cannot automatically use the uncertainty on the certificate as your uncertainty. Many thermometers, particularly low cost
thermistors, drift over their calibration cycle. I have been in
many laboratories that have thermometers with certificates
that say the uncertainty in the thermometer calibration is
0.01 ºC, but when examining the calibration history, I find
that the thermometer is adjusted by 0.05 ºC to 0.10 ºC or
more each time it is recalibrated. In general, the historical
variation in the thermometer is the acceptable uncertainty.
3. As in most “meets manufacturer’s specification” types of
calibrations, if the thermometer is not adjusted (see note 2
above), you can take the specification as a rectangular distribution.
To demonstrate the effects of thermally related sources of
uncertainty, the uncertainty budget for a single measurement is
analyzed. The first example is for a laboratory with only the
most basic knowledge of the environment, and succeeding
examples illustrate how the uncertainty can be lowered by
changing the level of temperature measurement and control.
The example is for the comparison of a test ring gage to a master
ring gage using a ULMM. [5] The master ring gage is calibrated
by an accredited laboratory with an uncertainty of 0.5 µm.
Uncertainty in CTE of the scale
u(αs) Vs S(ts)
Uncertainty in CTE of the test gage
u(αg) Vg [S(ts) + M20]
temperature of the scale
Uncertainty in CTE of the master gage
u(αm) Vm M20
tm
temperature of the master gage
Uncertainty in scale temperature
u(Vs) αs S(ts)
tg
temperature of the test gage
Uncertainty in test gage temperature
u(Vg) αg [S(ts) + M20]
Uncertainty in master gage temperature
u(Vm) αm M20
M20
calibrated length of the master gage
S(t s)
scale reading of the measurement instrument
Ta bl e 1. Quantities needed in order to make thermal
corrections to the measurement.
Tab le 2. Thermal error terms for the measurement.
MEASURE
|
33
TECHNICAL PAPERS
3. E x a m p l e 1 :
10 0 mm cus tom er ri ng gag e
c a li b ra te d usi ng a 100 mm m a ste r
ri ng o n a l on g ra ng e ULMM .
L abo r a to r y h as o n e t h er m o me te r
t o m o n i t o r ro o m .
Master Gage
100 mm Ring Gage
Uncertainty 500 nm (k = 2)
ULMM Specification
0.2 µm + 0.5 #10–6 L “Accuracy Specification” (Rectangular)
Test Gage Material: Steel
CTE = 12#10–6/ºC
Uncertainty 10 % (Rectangular)
Master Gage Material: Steel
CTE = 12 #10–6/ºC
The typical uncertainty budget for this
measurement equation is shown in Table 3.
Scale Material: Glass
CTE = 7#10–6/ºC
In addition:
•Room temperature variation is ±1 ºC.
Ta bl e 3. The values and uncertainties for the measurement in example 1.
Since we only have one thermometer
and it is used to determine some
sort of room average temperaStd
ture, we are hard pressed to say
Unc.
Uncertainty
Std. Unc.
Sensitivity
Standard
(µm)
/ Range
of Factor
Coeff.
Uncertainty
Source
Dist.
we know the temperature of
anything to better than 1 ºC. So,
Test Gage
1 ºC
Rect.
0.58 ºC
12#10–6 L/ºC
0.69
6.9#10–6 L
Temp. (Steel)
we will take the uncertainty in all
of the temperatures as a rectanMaster Gage
1 ºC
Rect.
0.58 ºC
12#10–6 L/ºC
0.69
6.9#10–6 L
Temp. (Steel)
gular distribution of ±1 ºC. Note
that if we were to measure the
Scale Temp.
1 ºC
Rect.
0.58 ºC
7#10–6 L/ºC
0.40
4.0#10–6 L
(Glass)
temperatures with the single
thermometer, the uncertainties
CTE (Scale)
Rect.
0.5 ºC
0.020
0.7#10–6/ºC
0.40#10–6/ºC
0.20#10–6 L
would be correlated and the
CTE
analysis would be more compliRect.
0.5 ºC
0.035
1.2#10–6/ºC
0.70#10–6/ºC
0.35#10–6 L
(Master Gage)
cated. Here, however, we are not
CTE (Test Gage) 1.2#10–6/ºC
Rect.
0.5 ºC
0.035
0.70#10–6/ºC
0.35#10–6 L
using the thermometer to
measure the actual temperatures
Length of
of the scale and gages, just to set
0.50 µm
Normal
0.250 µm
1
0.250 µm
0.25
Master Gage
limits on their variations.
Scale
We also have to estimate the
0.25 µm
Rect.
0.150 µm
1
0.150 µm
0.15
Specification
value of the temperature difference between the scale and gages
Combined Standard Uncertainty
1.10
and 20 °C. In a typical room the
Expanded Uncertainty (k = 2) 2.20
temperature change is roughly
linear with time, so the average
difference between 20 °C and
Ta bl e 4. Uncertainty calculation for example 1.
the scale and gages is 0.5 °C. It
can be argued that this is an over estimate because the gages and
ULMM act like low pass filters on the room air temperature, so
Standard
Std. Unc.
Ratio to
that the variations are more like sine waves than saw-tooth
Uncertainty
Squared
Largest
Source
waves. This calculation is too complex for most laboratories and
the changes are small compared to other sources of uncertainty.
Test Gage Temp. (Steel)
690 nm
476,100
1
Thus, we will use 0.5 °C. The calculation of the uncertainty comMaster Gage Temp. (Steel)
690 nm
476,100
1
ponents is shown in Table 4.
Scale Temp. (Glass)
400 nm
160,000
0.34
Let’s examine the biggest parts of this uncertainty. If we
compare them through their variances (standard uncertainty
CTE (Scale)
20 nm
400
0.001
squared), which is how they enter into the combined standard
uncertainty, we get the results shown in Table 5.
CTE (Master Gage)
35 nm
1,225
0.003
Since any source of uncertainty that is less than 1/3 to 1/4 of
CTE (Test Gage)
35 nm
1,225
0.003
the largest component does not significantly contribute to the
combined standard uncertainty, we see that there are three
Length of Master Gage
250 nm
62,500
0.131
sources of uncertainty that dominate our measurement (highScale Specification
150 nm
22,500
0.047
lighted). Since all of these are determined by the temperature
measurement, we can see that we need a better thermometer. To
SUM
1,200,050
reduce our uncertainty, the laboratory buys a portable thermometer that can be positioned on or near the gages when they are Table 5. Comparison of the relative sizes of each uncertainty
being measured; the new thermometer has a specification of 0.1°C. source for example 1 based on the ratio of variances.
34
|
MEASURE
www.ncsli.org
TECHNICAL PAPERS
Master Gage
100 mm Ring Gage
ULMM Specification
0.2 µm + 0.5 #10–6 L “Accuracy Specification” (Rectangular)
CTE = 12#10–6/ºC
4. E x a m p l e 2 :
10 0 m m te st ri ng ga g e c a li b ra te d
u si n g a 1 0 0 mm m as t er r in g o n a
l o ng ra n ge UL M M. La b o r a to ry
h a s p o r t a b l e t h e r m o m e t e r.
The typical uncertainty budget for this
measurement equation is shown in
–6
Table 6.
Scale Material: Glass
CTE = 7#10 /ºC
In addition:
• Test Gage and Master Gage temTa bl e 6. The values and uncertainties for the measurement in example 2.
peratures are measured and
corrected for using a digital
thermometer; its uncertainty
Std
specification is ±0.1 ºC.
Unc.
Uncertainty
Std. Unc.
Sensitivity
Standard
(µm)
/ Range
of Factor
Coeff.
Uncertainty
Source
Dist.
• Average temperature during
measurements = 20.45 ºC.
Test Gage
0.1 ºC
Rect.
0.058 ºC
12#10–6 L/ºC
0.7#10–6 L
0.069
Temp. (Steel)
• Scale temperature cannot
be measured.
Master Gage
0.1 ºC
Rect.
0.058 ºC
12#10–6 L/ºC
0.7#10–6 L
0.069
Temp. (Steel)
• Room temperature variaScale Temp.
tion is ±1 ºC.
1.0 ºC
Rect.
0.58 ºC
7#10–6 L/ºC
4.0#10–6 L
0.400
(Glass)
The calculation of the uncertainty components for this sitCTE (Scale)
0.7#10–6/ºC
Rect.
0.40#10–6/ºC
0.5 ºC
0.20#10–6 L
0.020
uation is shown in Table 7.
CTE
–6
–6
–6
Let’s examine the biggest
1.2#10 /ºC
Rect.
0.70#10 /ºC
0.45 ºC
0.31#10 L
0.031
(Master Gage)
components of this uncerCTE (Test Gage)
1.2#10–6/ºC
Rect.
0.70#10–6/ºC
0.45 ºC
0.31#10–6 L
0.031
tainty. If we compare them by
their variances (standard
Length of
0.50 µm
Normal
0.25 µm
1
0.250 µm
0.250
uncertainty squared), which is
Master Gage
how they enter into the comScale
0.25 µm
Rect.
0.15 µm
1
0.150 µm
0.150
bined standard uncertainty, we
Specification
get the results in Table 8.
We have now reduced our
Combined Standard Uncertainty 0.51
uncertainty quite a bit, but we
Expanded Uncertainty (k = 2) 1.02
still have an expanded uncertainty that is not very good
(1.02 µm). It is obvious where
we next need to improve our
process: the scale. One way is to make the scale out of a
material that has a very low CTE, like fused silica, or more
engineered materials like Zerodur or ULE Zero Expansion
Std. Unc.
Ratio to
Standard
Glass. Fused silica has a CTE of 0.5#10-6/ºC, which is conSquared
Largest
Uncertainty
Source
siderably less than glass. There are a number of engineered
Test Gage Temp. (Steel)
69 nm
4,761
0.03
materials with a CTE less than 0.1#10-6/ºC. All of these will
Master Gage Temp. (Steel)
69 nm
4,761
0.03
help. Another way is to put a thermometer on the scale, and
make corrections.
Scale Temp. (Glass)
400 nm
160,000
1.0
Another way that should help, but is problematic at times,
is
to have the scale be steel and measure steel parts. If the
CTE (Scale)
20 nm
400
0.003
scale and parts are the same temperature, and they usually
CTE (Master Gage)
31 nm
961
0.006
are closer in temperature to each other than the room variCTE (Test Gage)
31 nm
961
0.006
ation, you could get a lower differential thermal expansion
correction and therefore lower the uncertainty some. UnforLength of Master Gage
250 nm
62,500
0.39
tunately many instruments have the scale in a closed housing
and it is difficult to document the temperature difference.
Scale Specification
150 nm
22,500
0.14
CTE = 12 #10–6/ºC
SUM
256,844
Ta bl e 8. Comparison of the relative sizes of each uncertainty
source for example 2 based on the ratio of variances.
MEASURE
|
35
TECHNICAL PAPERS
5. E x a m p l e 3 :
1 0 0 mm t est ri n g ga ge
c al ib ra t e d u sin g a 1 0 0 mm
ma st e r r i ng o n a lo n g r a n g e
ULM M tha t has a lo w t her mal
c o e ff i c i e n t m a t e r ia l a s t h e
scal e. Labo r ato ry has po r tabl e
t h e r m o m e t e r.
Master Gage
100 mm Ring Gage
ULMM Specification
0.2 µm + 0.5 #10–6 L
“Accuracy Specification” (Rectangular)
CTE = 12#10–6/ºC
CTE = 12 #10–6/ºC
Scale Material: Low CTE
CTE = 0.1#10–6/ºC
The typical uncertainty budget for
this measurement equation is shown
in Table 9.
In addition:
• Test Gage and Master Gage
Std
temperatures are measured
Unc.
Uncertainty
Std. Unc.
Sensitivity
Standard
(µm)
/ Range
of Factor
Coeff.
Uncertainty
Source
Dist.
and corrected for using a
Test Gage
digital thermometer; uncer0.1 ºC
Rect.
0.058 ºC
12#10–6 L/ºC
0.69#10–6 L
69
Temp. (Steel)
tainty specification is ±0.1 ºC.
Master Gage
• Average temperature during
0.1 ºC
Rect.
0.058 ºC
12#10–6 L/ºC
0.69#10–6 L
69
Temp. (Steel)
measurements = 20.45 ºC.
Scale Temp.
• Scale temperature cannot be
1.0 ºC
Rect.
0.58 ºC
0.1#10–6 L/ºC
0.057#10–6 L
6
(Low CTE)
measured.
• Room temperature variation
CTE (Scale)
0.01#10–6/ºC Rect. 0.006#10–6/ºC
0.5 ºC
0.003#10–6 L
0.3
is ±1 ºC.
CTE
–6
–6
–6
1.2#10 /ºC
Rect.
0.70#10 /ºC
0.45 ºC
0.31#10 L
31
(Master Gage)
For our example, we will use a
fictional engineered material
CTE (Test Gage)
1.2#10–6/ºC
Rect.
0.70#10–6/ºC
0.45 ºC
0.31#10–6 L
31
that has a known thermal
Length of
expansion of 0.1 #10-6/ºC ±
500 nm
Normal
250 nm
1
50 nm
250
Master Gage
0.01#10–6/ºC. The calculation
Scale
of the uncertainty components
0.25 µm
Rect.
150 nm
1
150 nm
150
Specification
for this example is shown in
Table 10.
310
We now have only one large
Expanded Uncertainty (k = 2)
620
component we can change, the
uncertainty of the master ring.
Suppose we send the ring gage
Tabl e 10. Uncertainty calculation for example 3.
to NIST for calibration. At
NIST the ring is calibrated on our M48 coordinate measuring
corrected for using a digital thermometer; uncertainty specmachine in a laboratory that is temperature controlled to ±0.01
ification is ±0.1 ºC.
°C. The one directional repeatability of the M48 is less than 0.03
• Average temperature during measurements = 20.45 ºC.
µm, and our long term reproducibility studies shows a calibra• Scale temperature cannot be measured.
tion uncertainty for a 100 mm ring gage to be 0.12 µm.
• Room temperature variation is ±0.1 ºC.
With the development of long range instruments like the
The calculation of the uncertainty components for this example
ULMM, the need for matching the master gage to the test gage
is shown in Table 12.
is no longer necessary. This opens up the opportunity to use the
At this point, improvements become harder because there are
calibration services at NIST for laboratories that cannot afford
a number of thermal components of about the same size, and
the calibration of a whole gage block set. For nearly all size ring
the largest is from the ULMM. Fixing only one will not provide
gages, only a few masters are needed.
much improvement in the expanded uncertainty.
Basically, for most calibration laboratories, this is the end of
6. E x a m p l e 4 :
the road. If you look at industrial interlaboratory test data for ring
1 0 0 mm t e st rin g ga ge ca l ib ra t e d u s in g a 1 0 0 mm ma st e r
or plug gage measurements, you see things like those in Fig. 1. [6]
ri n g o n a lo n g r an ge U L M M wh ic h h a s a sc a le ma d e of a
The results in Fig. 1 are typical of a round robin for randomly
l o w C T E m a t e r i a l . L a b o r a t o r y h a s p o r t a b l e t h e r m o m e t e r.
selected calibration laboratories. In this round robin, there were
T h e m a s t e r r i n g i s c a l i b r a t ed a t N I ST.
two rings and the deviations are the differences from the NIST
calibration. The standard deviation between the laboratories
The typical uncertainty budget for this measurement equation
was, in all cases, around 0.4 µm. If we take this as a rough estiis shown in Table 11.
mate of the standard uncertainty of the group, we would get a
In addition:
k = 2 expanded uncertainty of around 0.8 µm. According to our
• Test Gage and Master Gage temperatures are measured and
36
|
MEASURE
www.ncsli.org
TECHNICAL PAPERS
Master Gage
100 mm Ring Gage
ULMM Specification
0.2 µm + 0.5 #10–6 L
“Accuracy Specification” (Rectangular)
CTE = 12#10–6/ºC
CTE = 12 #10–6/ºC
Scale Material: ULE-like
CTE = 0.1#10–6/ºC
uncertainty budget in Example 3,
this is about what we would expect
for measurements made in an
“average” environment (±1 ºC) with
decent thermometers (±0.1 ºC) using
a ULMM with a low expansion
material scale.
7. S u mma r y
Source
Uncertainty
/ Range
Dist.
Std. Unc.
of Factor
Sensitivity
Coeff.
Standard
Uncertainty
Std
Unc.
(µm)
Test Gage
Temp. (Steel)
0.1 ºC
Rect.
0.058 ºC
12#10–6 L/ºC
0.69#10–6 L
69
Master Gage
Temp. (Steel)
0.1 ºC
Rect.
0.058 ºC
12#10–6 L/ºC
0.69#10–6 L
69
Scale Temp.
(Low CTE)
0.1 ºC
Rect.
0.058 ºC
0.1#10–6 L/ºC
0.006#10–6 L
0.6
CTE (Scale)
0.01#10–6/ºC
Rect.
0.006#10–6/ºC
0.5 ºC
0.003#10–6 L
0.3
CTE
(Master Gage)
1.2#10–6/ºC
Rect.
0.70#10–6/ºC
0.45 ºC
0.31#10–6 L
31
CTE (Test Gage)
1.2#10–6/ºC
Rect.
0.70#10–6/ºC
0.45 ºC
0.31#10–6 L
31
Length of
Master Gage
120 nm
Normal
60 nm
1
60 nm
60
Scale
Specification
0.25 µm
Rect.
150 nm
1
150 nm
150
Thermal expansion is a critical part
of uncertainties for dimensional measurements. A fairly
simple analysis of the uncertainty components can be used
to decide how to rationally
allocate resources to obtain
adequate measurement performance.
8. Acknowledgment
I would like to thank Richard
Pettit and the anonymous
reviewer for pointing out a
remarkable number of numerical errors in this paper. The
paper is much stronger for their
efforts.
9. R ef er e n c e s
[1] Strangely, few metrology
books even mention thermal
expansion. A survey of over 50
Expanded Uncertainty (k = 2)
386
books on dimensional metrology or inspection revealed only
two with mentions of thermal
expansion, and those had only one paragraph each. The best
source for information on thermal expansion is ASME/ANSI
B89.6.2, Temperature and Humidity Environment for Dimen1.5
sional. It contains details on all of the concepts used in this paper.
[2] International Organization for Standardization (ISO), Guide to the
1
Expression of Uncertainty in Measurement, Geneva, Switzerland,
1993.
0.5
[3] American Society for Mechanical Engineering, ASME/ANSI
0
B89.1.9, Gage Blocks, New York, NY, 2002.
1
3
5
7
9 1 1 1 3 1 5 1 7 19 21 2 3 2 5 2 7 2 9 3 1 3 3
[4] There are a large number of reference books that list the CTE of
-0 . 5
various materials. Unfortunately the uncertainty is seldom
reported, and the CTEs reported in most sources are averages over
-1
large ranges of temperature, which increases the uncertainty at
20 °C. The 10 % uncertainty used in this paper is the consensus
-1 . 5
value used by experts in the field.
Lab
[5] How a 1D measuring machine came to be known as a “universal”
F i g u re 1. Deviation from the NIST calibration value for a 63.1825 mm
measuring machine is not known, but probably results from the
(2.4875 in.) ring gage as measured by 34 laboratories. The range of
the data is 2.0 µm, with a standard deviation of 0.4 µm.
fact that it can be used to measure both internal (ring) and external (plug) dimensions. A machine that measures in three dimensions is occasionally called a “universal measuring machine,” but
the most common term is “coordinate measuring machine.”
[6] Private correspondence.
193
Deviation from NIST (µm)
MEASURE
|
37
TECHNICAL PAPERS
1
Robert D. Moyer
A b s t r a c t : Scalar reflectometers afford a relatively inexpensive means to measure reflection coefficient magnitudes at
RF and microwave frequencies. Unfortunately, the measurements include errors that arise from vectorial imperfections
within the reflectometer hardware. While a vector network analyzer can help correct for the imperfections, a scalar
analyzer has only limited capability to do so, and there is often confusion about the extent of the corrections that can
be made. This paper provides a careful analysis of rf and microwave scalar reflectometers and discusses two common
ways to initialize them. The results (1) show the advantages, and remaining weaknesses, of the open/short method
for initializing a 4-port reflectometer, (2) demonstrate the irrelevance of detector reflection coefficients, (3) reveal the
advantages of 4-port over 3-port reflectometers, and (4) present expressions showing how the worst-case errors in
scalar reflectometer measurements vary, depending on the properties of both the reflectometer and the device under test.
1 . In t ro d u c t i o n
RF and microwave reflectometers are often used in conjunction
with scalar receivers to make moderately accurate reflection
coefficient measurements. Measurements using a scalar reflectometer are subject to errors that arise from vectorial imperfections in the reflectometer hardware. In this paper, a careful
analysis of both a 3-port and 4-port reflectometer is presented
which is then used to demonstrate the advantages of a 4-port
reflectometer over a 3-port reflectometer. The following five
sections discuss the theory involved in scalar reflectometer
measurements and includes two ways to initialize the reflectometer, as well as an analysis of the worst-case errors associated with a well-initialized scalar 4-port reflectometer.
Robert D. Moyer
Primary Standards Laboratory
Sandia National Laboratories
P.O. Box 5800
Albuquerque, NM 87185-0665 USA
38
|
MEASURE
2. R e p r e s e n t a t i o n o f R e f l e c t o m e t e r s a n d N o t a t i o n
Figures 1a and 1b show schematic representations of 4-port
reflectometers based on a directional coupler and a power splitter–directional bridge combination, respectively. For analysis
purposes, Fig. 1c shows the signal flow diagram representation
of a 4-port reflectometer. A signal generator, whose internal generator reflection coefficient is Γg, drives port 1 of the reflectometer. The initialization and unknown devices are connected
to port 2, the test port. The true reflection coefficient of any
device connected to test port 2 is denoted as z. The signal measured at port 3 is primarily proportional to the signal emergent
from test port 2 of the reflectometer (i.e., incident on the device
connected to port 2) while the signal measured at port 4 is primarily proportional to the signal reflected from the device connected to test port 2. Thus, the ratio of the latter signal to the
former signal is approximately proportional to the reflection
1 Sandia is a multi-program laboratory operated by Sandia Corporation,
a Lockheed Martin Company, for the United States Department of
Energy’s National Nuclear Security Administration under contract
DE-AC04-94AL85000.
www.ncsli.org
TECHNICAL PAPERS
z
3
4
2
1
b
a
2
2
s22
s23
s21
F i g u re 1a. Directional coupler.
bg
s32
b
a1
3
s31
1
Γg
s11
b1
3
s34
s42
s24
s12
ρ3 →
s33
a
s13
3
4
s14
s43
s41
s44
2
F i g u re 1b. Splitter – directional bridge.
a4
b4
ρ4
↓
F i g u re 1c. 4-port signal flow diagram.
coefficient of the device connected to port 2. The ratio of the
traveling wave voltage emergent from port 4, b4, to the traveling wave voltage emergent from the generator, bg, may be
z ( S 21 S 42 − S 22 S 41 ) + S 41
b
written by inspection of the signal flow graph (see Section 2.18
,
(3)
wm = 4 =
b3 z ( S 21 S 32 − S 22 S 31 ) + S 31
of Ref. [1]) as:
 S S − S or
S 41
S 41 
22
z  21 42
+
1−
Γ
S
1−
Γ S
1−
z
S
+
S
zS
S
b4


g
11
41 (
22 )
21
42
S 42+−e S 22 S 41 ) S 31 + S 41 S 31 qz + r
b 4g 11z ( S=21hz
=
=
(1)
, (4)
w
=
=
bg
1− Γg S11 − zS 22 + Γg S11 z S 22 − Γg S 21 zS12
 Γg S11 S 22 − Γg S 21 S12m − Sb22 =
fz +1

sz +1
S S 
3
z
 +1 −z  S 22 − 21 32  +1
1− Γg S11


S 31 

S S −S S 
S 41
21
42
22
41
z
+
− z S 22 ) + S 21 zS 42
 1− Γg S11
 1− Γg S11
hz + e ,
where q, r and s are complex functions of the S-parameters of
=
=
the reflectometer as indicated in Eq. (4).
+ Γg S11 z S 22 − Γg S 21 zS12
 Γg S11 S 22 − Γg S 21 S12 − S 22 
fz +1
z
 +1
For 3-port reflectometers, one uses Eq. (1) for the measured
1− Γg S11


ratio because, using the notation of Fig. 1, port 3 is not present;
for 4-port reflectometers, one uses Eq. (4).
where:
h, e and f are functions of Γg and the S-parameters of the reflec3 . A G o o d Wa y t o I n i t i a l i z e S c a l a r R e f l e c t o m e t e r s
tometer as suggested in Eq. (1);
ρ3 and ρ4 (in Fig. 1c) are the reflection coefficients of the detecIf a vector receiver is available to measure the complex ratios
tors on ports 3 and 4, respectively; and
and if the complex S-parameters of the reflectometer are
known, then very accurate reflection coefficient measurements
ρ3 and ρ4 were set to 0 during the derivation of Eq. (1) in anticcan be made. It is, however, much less expensive to use scalar
ipation of the result of section 4. (There it is shown that the
sensors, which respond only to amplitudes, on a reflectometer.
values of ρ3 and ρ4 are arbitrary when the reflectometer is
Unfortunately, accuracy decreases when scalar detectors are
initialized by the preferred technique described in section
used on a 4-port reflectometer. This section discusses a way to
3.) Equations (1) through (4) would be much more cummitigate the loss of accuracy through use of a well-known inibersome if ρ3 ) 0 and ρ4 ) 0.
tialization procedure.
Similarly, replacing 4’s by 3’s in Eq. (1):
Equations (1) and (4) show that both b4/bg (measured on a
3-port
reflectometer) and b4/b3 (measured on a 4-port reflecS S −S S 
S 31
22
31
z  21 32
+
tometer)
are given by linear fractional expressions. Because of
b3
 1− Γg S11
 1− Γg S11
.
(2)
this,
the
discussion
in this section applies to both 3- and 4-port
=
bg
 Γg S11 S 22 − Γg S 21 S12 − S 22 
reflectometers.
For
convenience,
however, we will use the notaz
 +1
1−
Γ
S
tion
for
the
4-port
reflectometer.


g
11
If one simply connects a device under test (DUT) to the test
port of a scalar reflectometer and measures it, the scalar reflecTaking the ratio of the preceding two equations gives the meastometer produces (at least internally) a value, |wm|, which is
ured ratio, wm, of the emergent traveling wave voltage from
port 4 to that at port 3 as:
assumed to obey the simple model:
MEASURE
|
39
TECHNICAL PAPERS
wm = Q wt ,
| wm |
where wt is the true reflection coefficient of the DUT and Q is
a scalar constant determined by the scalar reflectometer hardware. In reality, |wm| obeys the more complex model of Eq. (5),
which is merely the magnitude of Eq. (4):
| wm | =
| b 4 | | qz + r |
.
=
| b3 | | s z +1|
(5)
or
| wm |
ws wo
| q z + r | | se − j φ +1| | −se − j φ +1|
,
| sz +1| | qe − j φ + r | | −qe − j φ + r |
=
| q z + r | | −s 2 e − j 2 φ +1|
,
| sz +1| | −q 2 e − j 2 φ + r 2 |
or
| −s 2 e − j 2 φ +1|
If r = s = 0 (as in a perfect reflectometer), Eq. (5) collapses to
the simple model. Nevertheless, even in a perfect reflectometer,
the value of q differs significantly from unity as shown by the
coefficient of z in the numerator of Eq. (4). Equation (5) therefore provides a poor estimate of |wt|; the reflectometer needs to
be “initialized” in order to obtain a better estimate of |wt|.
First, a poor way to initialize the reflectometer is described in
the following three steps:
Step 1A: Measure an open circuit, i.e., z = e–jφ, at test port 2
which gives:
| qe − j φ + r | .
| wo | =
| se − j φ +1|
ws wo
=
(6)
| wm |
| −e − j 2 φ + ( r q ) |
2
=
| −s 2 e − j 2 φ +1|
| −e − j 2 φ + ( r q ) |
| sz +1|
ws wo
or
z + (r q )
| wm |
ws wo
=
| q'' z + r" | .
| s z +1|
2
. (10)
(11)
where q” is the coefficient of z in the numerator of Eq. (10)
and r” is the final term in that numerator. From Eq. (4), it
follows that:
S 41
S 41
r
,
=
≈
q S 21 S 42 − S 22 S 41 S 21 S 42
(12)
where the approximation holds because |S22S41| << |S21S42|
Step 2A: Divide Eq. (5), which results from measuring the
unknown device, by Eq. (6) to obtain (after re-arranging):
for both circuits shown in Fig. 1. Furthermore, |r/q|2 ;
− jφ
|d/S
|2 << 1 since the coupler or directional bridge will have
| se +1|
| se − j φ +1|
z − jφ
+ ( r q ) − j φ 21
− jφ
high
directivity,
i.e. very small |d| = |S41/S42|, if it is used in a
| e + (r q ) |
| e + (r q ) |
| wm | | q z + r | | se +1|
(7)
=
=
reflectometer
application.
Also, as previously mentioned, |s|2
| sz +1|
| sz +1| | qe − j φ + r |
| wo |
<<1 so the coefficient of z in the numerator of Eq. (10)
reduces to a value very near unity.
− jφ
− jφ
| se +1|
| se +1|
z − jφ
+ (r q ) − j φ
Equations (10) and (11) therefore show that
| e + (r q ) |
| e + (r q ) |
| wm | | q z + r | | se − j φ +1|
,
=
=
| wm |
| sz +1|
| sz +1| | qe − j φ + r |
| wo |
ws wo
or
| wm | | zq'+ r ' | .
=
| wo | | s z +1|
(8)
Step 3A: Equation (8) shows that the estimator |wm/wo|
equals |z| if r' = s = 0 and |q'| = 1. Unfortunately, r' and s are
often large enough to drive |q'| away from unity because q'
has first order dependence on r and s as seen in Eq. (7).
A preferred, well-known, procedure for initialization is given
in step 1B through step 3B which follow:
Step 1B: Measure the unknown and open circuit to get Eqs.
(5) and (6), respectively, as shown above.
Step 2B: Measure an offset short circuit, i.e., z = e–j(π+φ ),
which is π radians out of phase with the open circuit, to get:
| ws | =
| qe − j (π + φ ) + r | | −qe − j φ + r | .
=
| se − j (π + φ ) +1| | −se − j φ +1|
(9)
Note: A short/open pair which very closely provides the π
radian phase difference is available from most manufacturers
of scalar reflectometers. See Appendix A.
Step 3B: Divide Eq. (5) by the square root of the product of
Eqs. (6) and (9) to obtain:
40
|
MEASURE
is a better estimator because q" has only second order dependence on the small, but inevitably present, r and s parameters!
|q"| will therefore more closely approximate unity than does |q'|
as defined by Eqs. (7) and (8). Steps 1B through 3B therefore
define a preferred technique for initializing a scalar reflectometer.
The open-short initialization technique gives a value of q"
that is very near unity. While this is an advantage, it is unfortunately the only benefit that results from use of the open-short
initialization technique. First, notice that the denominators of
the right-hand-sides of Eqs. (5), (8) and (11) are identical,
namely |s z + 1|; hence the harmful effects of s are the same in
all three equations. Engen [2] has shown that Γge = –s is the
effective generator reflection coefficient of the 4-port reflectometer. Results from Eq. (11) therefore contain unmitigated
errors due to Γge. Equation (4) may be rewritten as:

r
qz + 
q

.
wm =
sz +1
(4a)
Clearly, the numerator of the expression for wm includes an
undesirable component, r/q, which is a function only of reflectometer properties and is independent of z. Again, from Eq. (12),
www.ncsli.org
TECHNICAL PAPERS
r
=
q
S 41
S 42
S 21 − S 22
,
S 41
S 42
(12a)
where S41/S42 is the directivity as observed at port 4 of the
reflectometer. Equations (7) and (8) reveal that r'/q' = r/q while
Eqs. (10 and (11) show that r"/q" = r/q. This shows that the
preferred initialization does nothing to mitigate the effects of
the directivity of the reflectometer. Thus, measurements of
unknowns will still be in error due to (a) multiple reflections
between the unknown and the effective generator reflection
coefficient of the reflectometer, and (b) the directivity of the
reflectometer.
4. I r re l e v a n c e o f D e t e c t o r R e f l e c t i o n C o e ff i c i e n t s
Equations (1) through (4) involve the traveling wave voltages,
while most detectors respond to power or the voltage squared.
Furthermore, detectors always reflect a small portion of any
signal incident upon them. However, the reflection coefficients
(ρ3 and ρ4 in Fig. 1c) of the detectors do not influence the
results from a scalar reflectometer that is initialized by the preferred technique described in the preceding section. Again, this
is true for both 3- and 4-port reflectometers as shown in the following paragraphs.
In practice, power, rather than bix, (i = 3 or 4; x = ‘m’ for
unknown being measured, ‘s’ for short and ‘o’ for open measured
during initialization) is measured at each sidearm port of a scalar
reflectometer. As shown by Eq. (2.37) of Ref. [1], the power
incident on detector 4 in Fig. 1c is given by |b4x|2 /Z0. Similarly,
the power reflected from detector 4 is |b4xρ4|2 /Z0. Therefore, the
power absorbed by detector 4 is P4x = |b4x|2 (1 – ρ42) / Z0 where:
Pix is the power absorbed by detector i when voltage wave bix
is incident on it;
Z0 is the (real) characteristic impedance of the system; and
ρi is the magnitude of the reflection coefficient of detector i.
Consequently,
P4x Z 0
b4x =
.
1 − ρ 42
The ratio
(
)
)
(
)
P4 x Z 0 1 − ρ 32
P4x 1 − ρ 32
b 4x
=
=
2
b3x
1 − ρ 4 P3x Z 0
1 − ρ 42 P3x
(
(
)
(13)
therefore follows.
Using Eq. (13) and referring to Eqs. (4) and (11), the preferred estimator for a 4-port scalar reflectometer is given by:
b4 m
b3 m
b4 s b4o
b3 s b3o
(
P4 m 1 − ρ 32
)
(1 − ρ ) P
=
P (1 − ρ ) P (1 − ρ )
(1 − ρ ) P (1 − ρ ) P
2
=
4
3m
2
4s
3
2
4
2
4o
3
2
3s
4
3o
P4 m
P3 m
P4 s
P3 s
.
P4 o
P3o
It is important to notice that parameters q, r and s in Eq. (4) are
entirely independent of Γg, the reflection coefficient of the generator which drives the 4-port reflectometer. It therefore follows
that the preferred estimator, see Eq. (11), for a 4-port scalar
reflectometer is also independent of Γg. On the other hand, if a
3-port reflectometer is used, one must work with Eq. (1)
because port 3 does not exist. In this case, the parameter f in Eq.
(1) has three terms which depend on Γg ; the ΓgS21S12 term, in
particular, can be large enough to cause significant errors in the
measured reflection coefficient. Furthermore, it was pointed out
in the last paragraph of section 3 that the parameter s (for a 4port reflectometer) is not diminished by use of the preferred initialization procedure. It is easily shown that the same is true of
the parameter f for the 3-port reflectometer. The parameter f,
and therefore the error in measurements from a 3-port reflectometer, changes whenever Γg (the generator reflection coefficient of the generator driving the 3-port reflectometer) changes.
Based on these considerations, two significant advantages of
the 4-port reflectometer are:
1. The performance of the 4-port reflectometer is independent
of Γg, the reflection coefficient of the generator which
drives it, while a 3-port reflectometer’s performance
depends on the reflection coefficient of its driving generator.
2. If b3 and b4 are measured simultaneously, the 4-port reflectometer is immune to any changes in generator level that
may occur while carrying out the initialization and DUT
measurements. On the other hand, the 3-port reflectometer
provides no mechanism for eliminating the effects of such
changes.
6. Wo r s t - C a s e E r ro r s o n M e a s u r e m e n t s f ro m a
We l l - I n i t i a l i z e d S c a l a r 4 - P o r t R e f l e c t o m e t e r
Equation (11), as a real equation, specifies the preferred estimator from a scalar reflectometer; the magnitude bars are required
because the scalar detectors cannot sense vector information.
However, the error mechanisms in the scalar reflectometer are
vector in nature. In order to characterize the errors in the scalar
estimator, it is necessary to analyze the vector relationships
obtained by removing the magnitude restraints from Eq. (11).
Thus:
wm
q'' z + r "
(15)
=
=w .
sz +1
ws wo
Equation (15) shows that the vector estimator w is a complex
bilinear transformation of z, the true vector reflection coefficient of the device being measured.
Inverting Eq. (15) gives:
(14)
This shows that the preferred estimator is independent of both
detector reflection coefficients.
5. A d v a n t a g e s o f a 4 - P o r t R e f l e c t o m e t e r
Ov e r a 3-P o r t R e f l e ct o me t e r
z=
w −r"
a"w + b " .
=
−s w + q"
cw +1
(16)
Therefore, if the locus of w is a circle at the origin of the
complex plane, the locus of z will be another circle of radius R1
whose center lies at the tip of a vector C1. According to Eqs.
MEASURE
|
41
TECHNICAL PAPERS
C
z locus
1
R
1
|w|
w locus
F i g u re 2. A possible relationship between w and z loci.
(2.153) and (2.154) from Kerns and Beatty [1], R1 and C1 are
given by:
| a"− b "c || w |
1− |cw| 2
(2.153)
b "− a"c* | w | 2 , respectively.
1− | cw |2
(2.154)
R1 =
and
C1 =
Knowing that |a"| ~ 1, |b"| < 1 and |c| < 1 for a well-initialized
4-port scalar reflectometer, Eq. (2.153) shows that R1 will be
approximately equal to |w| while Eq. (2.154) shows that |C1| will
very likely exceed R1 for small values of |w|. Figure 2 depicts a
possible relationship between the locus of w, which is centered
at the origin, and the locus of z, which is centered at the tip of C1.
The worst-case error at a given frequency ν, WCEν, in a scalar
reflectometer measurement, |w|, is the largest difference
between |w| and |z|. It is given by the expressions (see Appendix B):
WCEν = |C1| + R1 – |w| if R1 > |w| or |C1| > |w|.
(17a)
If |w| > R1 > |C1| then WCEν = |C1| – R1 + |w|.
(17b)
which result from the specified values of a", b" c and |w|. The
sixth column identifies the appropriate equation, as determined
by the relative values of |w|, R1 and |C1|, for computing WCEν.
Finally, the computed value of WCEν appears in the seventh
column.
Notice that WCEν decreases as |w| increases when the phase
of c is 0 in cases 1 and 2. On the other hand, WCEν increases
with increasing |w| when the phase of c is 180°.
The expressions for worst-case error, WCEν, given in Eqs.
(17a,b,c) followed quite naturally from the physical operation
of a 4-port scalar reflectometer. To conform with current
popular practice, one could try to re-cast these expressions into
“standard uncertainties” or “expanded uncertainties” as recommended by the U.S. Guide to the Expression of Uncertainty in
Measurement (or Guide). [3] It appears, however, that this
would not be appropriate. Paragraph 3.2.4 of the Guide states:
“It is assumed that the result of a measurement has been corrected for all recognized significant systematic effects and that
every effort has been made to identify such effects.”
There are three sources of systematic effects in the expression
for w from a scalar reflectometer, namely the q”, r” and s parameters in Eq. (15). These are recognized and significant quantities and produce systematic effects that depend on the value of
the device measured. Equations (2.153) and (2.154) lead to
Fig. 2 which graphically shows the systematic effects of the
three error sources. While it is possible to correct for these systematic effects in vector network analyzer systems, it is not possible in scalar systems.
The Guide does offer an example (in section F.2.4.5) of a way
to incorporate “a single mean correction” into the combined
variance to accommodate situations where it is impossible or
undesirable to eliminate known systematic effects. It appears,
however, that such a procedure would produce an uncertainty
statement that is not related to the physical processes in the
reflectometer which cause the errors, as shown in Fig. 2. Thus
the uncertainty statement is of no added value.
7. C o n clu s io n s
If |w| > |C1| > R1 then WCEν = |C1| – R1 – |w|.
(17c)
Some important remarks about application of Eqs. (17a,b,c) are
in order. If one uses the complex values (at a given frequency ν)
of a", b" and c to compute the values of R1 and |C1|, then using
the appropriate equation of (17) gives WCEν the worst-case
error with respect to all possible phases of w at frequency ν.
However, WCEν is not worst-case with respect to all possible
phases of a", b" and c at frequency ν.
It is important to recognize that, at a particular frequency,
WCEν may either increase or decrease as |w| increases. This is
demonstrated in Table 1 where values of WCEν are computed
for a scalar reflectometer having typical values a" = 1, b" = 0.01
and |c| = 0.1. In the first two cases considered, the phase of c is
0°; in the final two cases, the phase of c is 180°. Cases 1 and 3
show the situation when the measured reflection coefficient is
|w| = 0.1 while cases 2 and 4 apply when |w| = 0.3. The fourth
and fifth columns give calculated (see Eqs. 2.153 and 2.154)
values for radius R1 and magnitude of vector C1, respectively,
42
|
MEASURE
The advantage of using an open/short method for initializing the
scalar reflectometer has been quantified; it produces a calibration constant, q", that is very near to unity. On the other hand,
the directivity and effective generator reflection coefficient of
the reflectometer remain as error sources. It was demonstrated
that the reflection coefficient(s) of the detector(s) have no effect
on the performance of either 3-port or 4-port reflectometers. It
was also shown that the performance of a 4-port scalar reflec-
Case
c
|w|
R1
|C1|
1
0.1 0.1
Equation WCEν
0.09991 0.009
(17b)
0.009
2
0.1 0.3
0.29997 0.001
(17b)
0.001
3
–0.1 0.1
0.10011 0.011
(17a)
0.011
4
–0.1 0.3
0.30057 0.019
(17a)
0.020
Ta bl e 1. Dependence of WCEν on |w| and phase of c.
www.ncsli.org
TECHNICAL PAPERS
terms of magnitude |εs| and |ε(r/q)| have been added to the
numerator and denominator, respectively, under the radical.
Hence, if the difference between the short and open phase
angles is (π + ε) radians, the effect is to introduce two secondorder errors, each proportional to ε, into the results obtained for
the ideal case when the phase angle difference is exactly π
radians.
tometer is entirely independent of both the generator level instability (if the detector outputs 3 and 4 are measured simultaneously)
and the generator reflection coefficient of its driving generator.
A 3-port reflectometer is vulnerable to both. The worst-case
error of a scalar reflectometer measurement depends on the
scattering parameters of the reflectometer and the value of the
reflection coefficient, |w|, of the measured device. Three different expressions for worst-case error from a 4-port reflectometer, depending on the scattering parameters of the reflectometer
and |w|, were given.
9. A ppendix B
This appendix justifies the expressions for WCE given in Eqs.
(17a,b,c), subject to the associated conditions. It also develops
the logical conditions for use of the different expressions for
WCE. Figure 2 in the body of the report depicted a possible relationship between the loci of w and z. For the purposes of this
appendix and convenience, one need consider only the situation
when arg(C1) = 0 as shown in Fig. B1.
For the sake of even more brevity, all the necessary information in Fig. B1 can be condensed into Fig. B2.
The caption of the figure specifies the size order of |C1|, R1
and |w|. In the figure, the points where the loci intersect the real
axis are identified as is the tip of the vector C1. The center part
of Fig. B2 graphically shows the worst-case error, WCE, as the
distance between (1) the point, zr, on the z locus that is most
remote from the w locus and (2) the point on the w locus that
is nearest to zr. In the right side of the figure, the corresponding
R 1 WCE is given. Figures B3 through B7
algebraic expression for
show corresponding information for the other five order possibilities.
In Figs. B2 through B5, the same expression is used for the
C1
WCE while different
expressions are required in Figs. B6 and
B7. This situation is logically summarized in the Karnaugh map
of Fig. B8. Notation used is asz follows:
X identifies a “don’t
locus
8. A ppendix A
The coefficient q", as defined by Eqs. (10) and (11), was
derived using the assumption stated in Step 2B of section 3, i.e.,
that the phase angle of the short used during initialization differed from the phase angle of the open by exactly π radians.
Unfortunately, in practice, the two phase angles will nearly
always differ by (π + ε) radians, where ε is some small value.
Broadband measurements of four different short/open pairs in
three different connector types from two different manufacturers revealed a maximum deviation of | ε | = 0.2 radian. The following development examines the effects such an ε deviation.
Let the reflection coefficient, z’, of the “offset” short be z' =
e–j(π+φ+ε). Equation (9) then takes the form:
−qe − j ϕ e − j ε + r
ws' =
.
−se − j ϕ e − j ε +1|w|
(9')
Divide Eq. (5) by the square root of the product of Eqs. (6) and
(9') to obtain:
| wm |
ws' wo
− jϕ
+1| | −se − j ϕ e − j ε +1|
| q z + r | | se
w locus
,
− jϕ
| sz +1| | qe + r | | −qe − j ϕ e − j ε + r |
=
(
(
)
)
(
(
)
)
or, after manipulation, Eq. (10') which is:
| −s 2 e − j 2 ϕ e − j ε + se − j ϕ 1− e − j ε +1|
| −s 2 e − j 2 ϕ e − j ε + se − j ϕ 1− e − j ε +1|
WCE
WCE = |C 1| + R1 - |w|
z
+
r
q
2 − j 2ϕ − j ε
− jϕ
− jε
2
2 − j 2ϕ − j ε
− jϕ
− jε
|
−q
|
−q
e
e
+
rqe
1−
e
+
r
|
e
e
+
rqe
1−
e
+r2 |
| wm |
(10')
=
| sz +1|
R
ws' wo
1
|w|
q
| wm |
'
(
)
(1− e ) + r
| −s 2 e − j 2 ϕ e − j ε + se − j ϕ 1− e − j ε +1|
| −q e
2
− j 2ϕ
e
− jε
+ rqe
− jϕ
=
− jε
2
|
(
)
(1− e ) + r
C1
| −s 2 e − j 2 ϕ e − j ε + se − j ϕ 1− e − j ε +1|
z +r
| −q e
2
− j 2ϕ
e
− jε
+ rqe
− jϕ
− jε
2
|
.
w locus
| sz +1|
ws wo
zF i g u re B1. A possible
C
relationshipzbetween
w and z loci when
r
1
arg(C1) = 0.
| −e − j 2 ϕ + jε ( r q ) e − j ϕ + ( r q ) |
(10")
2
| wm |
ws' wo
| −s 2 e − j 2 ϕ + jεse − j ϕ +1|
'
ws wo
WCE = |C 1| + R1 - |w|
WCE
wε). If ε << 1, cos(ε) ; 1 and
w
Observe that e–jε = cos( ε) – j sin(
sin(ε) = ε, so Eq. (10') simplifies to Eq. (10"), which is:
| −s 2 e − j 2 ϕ + jεse − j ϕ +1|
| wm |
z locus
≈
| −e
− j 2ϕ
+ jε ( r q ) e
− jϕ
+ (r q ) |
2
≈
z + (r q )
| −e
+ jε ( r q ) e
− jϕ
| sz +1|
w
| −s 2 e − j 2 ϕ + jεse − j ϕ +1|
− j 2ϕ
z + (r q )
+ (r q ) |
2
| −s 2 e − j 2 ϕ + jεse − j ϕ +1|
| −e − j 2 ϕ + jε ( r q ) e − j ϕ + ( r q ) |
2
WCE = |C 1| + R1 - |w|
WCE
w
z
C
1
z
r
.
| sz +1|
WCE = |C 1| + R1 - |w|
WCE
Equation (10") is the same as Eq. (10) except that second-order
F i g u re B2. Condition: |C1| > R1> |w|.
MEASURE
|
43
w
w
z
TECHNICAL PAPERS
z
C1
r
WCE = |C 1| + R1 - |w|
WCE
w
w
z
z
C1
r
R1 > | C1 |
________________
2 - B6
WCE = |C 1| + R1 - |w|
WCE
X
wC
z w
z
1
F i g u re B3. Condition:
| C 1| > |w | > r R 1.
1 - B5
X
3 - B7
1 - B4
1 - B2
1 - B3
| C1| > | w|
________________
R 1 > |w|
WCE
z w
wC
WCE = |C1 | + R1 - |w|
z
1
r
WCE
WCE = |C1 | + R1 - |w|
wC
z w
z
1
F i g u re B4. Condition:
R1 > |C1| >r |w|.
WCE
z
w
C1 w
WCE = |C1 | + R1 - |w|
z
F i g u re B8. Karnaugh map of expressions for WCE.
care” or impossible condition. The expression for WCE used in
Figs. B2 through B5 is identified as expression “1” and is
denoted by a leading “1 –“. The expression for WCE used in Fig.
B6 is identified as expression “2” and is denoted by a leading
“2 –“. The expression for WCE used in Fig. B7 is identified as
expression “3” and is denoted by a leading “3 –“. The figure
where each expression appears is indicated as a trailing (for
example) “– B7”.
The map indicates that:
• Expression 1 should be used if R1 > |w| or |C1| > |w|.
• Expression 2 should be used if |w| > R1 and R1 > |C1|, i.e. if
|w| > R1 > |C1|.
r
• Expression 3 should be used if |w| > |C1| and |C1| > R1, i.e. if
|w| > |C1| > R1.
WCE = |C1 | + R1 - |w|
WCE
w
z
C
F i g u re B5. Condition: R1 > |rw| > |C1|. 1
z
r
WCE
w
C
z
1
w
WCE = |C 1| - R1 + |w|
w
Z r C1 Z
F i g u re B6. Condition: |w| > R1 > |C1|.
Zr
C
1
Z
WCE
11. Re f e re n c e s
[1] D.M. Kerns and R.W. Beatty, Basic Theory of Waveguide Junctions
and Introductory Microwave Network Analysis, Pergamon Press,
Oxford, 1967.
[2] G.F. Engen, “Amplitude stabilization of a microwave signal
source,” IRE Trans. Microwave Theory and Technique, vol. MTT6, pp. 202-206, 1958.
[3] “U.S. Guide to the Expression of Uncertainty in Measurement,”
ANSI/NCSL Z540-2-1997, National Conference of Standards Laboratories International, Boulder CO.
WCE = |C | - R1 - |w|
WCE
w
10. Ac kn ow le dg e me nt s
The measurements made by James A. Woods and cited in
Appendix A are gratefully acknowledged.
WCE = |C 1| - R1 + |w|
WCE
w
w
z
1
w
WCE = |C | - R1 - |w|
1
F i g u re B7. Condition: |w| > |C1| > R1.
44
|
MEASURE
www.ncsli.org
MEASURE
|
45
TECHNICAL PAPERS
A Direct Comparison System
for Measuring Radio Frequency Power
(100 kHz to 18 GHz)
R onal d G i nl ey
A b s t r a c t : A direct comparison power measurement system has been developed to measure power sensor effective effi-
ciency in the 100 kHz to 18 GHz frequency range. This system is capable of measuring thermistor and thermoelectric based power sensors. Several problems needed to be addressed in the development of the system, including RF
leakage from the power sensors and its effect on system electronics, the sensitivity of the power meter and digital volt
meter to extraneous signals, and the effect of compensation beads, if there were any, in the sensors. This article covers
these problems, provides a discussion of the system design, presents the uncertainty analysis for the system, and finally
compares measurement results to measurements made using the NIST 0.05 GHz to 50 GHz system and the voltage/
impedance technique.
1 . I n t ro d u c t i o n
A system’s output power level is frequently the critical factor in the design,
and ultimately in the purchase and performance, of almost all radio frequency
and microwave equipment. [1] Measurements of microwave power are important for a wide array of electronic
devices. Equipment, whose characteristics and performance are determined by
microwave power measurements, can be
found in the areas of communications,
aerospace, navigation, surveillance, manufacturing, medical, and consumer electronics. Power measurements are also
the foundation for many other
microwave measurements such as attenuation and impedance.
This paper describes a new system for
measuring microwave power at the
National Institute of Standards and Technology (NIST). This system covers the
frequency range of 100 kHz to 18 GHz
and approximately 1 mW to 10 mW. This
is the first NIST system that is capable of
covering this entire frequency range
with one system. The system is also
Resistive
Power
Splitter
100 kHz to
20 GHz
Signal
Generator
46
|
MEASURE
Monitor
Power
Sensor
Power
Meter
DVM
Power
Meter
DVM
F i g u re 1. Block Diagram of the 100 kHz to 18 GHz direct comparison system.
capable of measuring thermoelectric type
of standards across the full 100 kHz to
18 GHz range. This system is similar to
the NIST 0.05 to 50 GHz system, but we
were previously unable to measure
devices below 50 MHz. The basic theory
and design of the system will be covered.
Some of the problems encountered in the
development of the system will be briefly
described. The uncertainty components
for the system will be briefly discussed.
Finally the new system and existing
NIST systems will be compared.
Ronald Ginley
Electromagnetics Division
and Technology1
325 Broadway
Standard
or DUT
2 . Sy st em De si gn an d Th e ory
Figure 1 shows a block diagram of the
system. Overall the system is very simple.
A signal generator sends an RF signal
into a resistive power splitter that then
splits the signal between a monitor
detector and either the calibration standard detector or the device under test
(DUT). The detectors are connected to
power meters whose output is connected
to a digital volt meter (DVM) in the case
of a thermistor type detector, or whose
output is read directly by a connected
computer through an instrument interface bus for a thermoelectric type of
detector. This type of system design is
not new and has been used at NIST
before. [2, 3]
The calibration and measurement
processes are detailed in reference. [2] A
synopsis of the calibration and measurement process follows. To calibrate the
1 U.S. Government work is not protected by U.S. copyright.
www.ncsli.org
TECHNICAL PAPERS
Ka =
Pdc–std
η
std
PM–std M gl–std
,
(1)
Effective Efficiency
where:
Pdc-std is the power read from the calibration standard;
ηstd is the known effective efficiency of
the calibration standard;
PM-std is the power read from the monitor
detector during the calibration;
and
Mgl-std is the mismatch factor from the
reflection coefficients of the standard and the splitter.
The process for measuring the ηe of
the DUT is the reverse of the calibration
process. From the power readings at the
monitor detector and the DUT, the
0.995
reflection coefficient of the DUT, the
reflection coefficient
looking into the test
0.990
port, and Ka, the η e of the DUT can be
determined. Note0.985
that all of the power
readings are used in ratios (standard or
0.980
DUT to the monitor
detector) and are
never used as an0.975
absolute power value.
By measuring the powers always in ratio,
any drift of the signal
0.970 power amplitude is
negated. The measurement process is
represented by: 0.965
η
=
DUT
where
0.960
Pdc–DUT
Ka P
0.955
M–DUT
Mgl–DUT
(2)
0.950
ηDUT is the effective efficiency of the
0.945
DUT;
0
5
Pdc-DUT is the power read from the DUT;
PM-DUT is the power read from the
monitor detector during the
DUT measurement; and
Mgl-DUT is the mismatch factor from the
reflection coefficients of the
DUT and the splitter.
Several reflection coefficients are used
during the measurements. This is necessary to correct for the impedance mismatches between various components of
the system (Mgl-std and Mgl-DUT). The
reflection coefficient of the standard and
the DUT are measured directly on vector
network analyzers (VNAs), and the
reflection coefficient looking into the test
port of the splitter is determined with a
modified VNA calibration scheme. [4]
To summarize the operation of the
system, a standard with known effective
efficiency is used to determine a value
that relates the power at the test port to
the power measured in the monitor
detector. This value is then used to determine the effective efficiency of an
unknown device from the power readings of the unknown device and the
monitor detector. The impedance mismatches are corrected for throughout the
process.
3 . C a l i b r a t i o n S ta n d a rd s
It is not possible to cover the entire
working frequency range of this system
with one standard. Two standards are
used. The first covers the 100 kHz to 50
MHz range. This standard is calibrated
using voltage and impedance measurements. From the voltage and the equivalent parallel input resistance, the
effective efficiency of the detector can be
determined. [5] The 1σ uncertainty in
the calibration of the effective efficiency
of the standard ranges from 0.00046 to
0.00133 as the frequency goes from
100 kHz to 50 MHz. The second type of
standard, used from 50 MHz to 18 GHz,
is evaluated in the NIST microcalorimeter. [6] The 1σ uncertainty for this
evaluation ranges from 0.0012 to 0.0021
across the frequency range.
4. P ro b l e m s E n c o u n t e re d i n t h e
D e v e l op me nt o f the S y s te m
4 .1 C ompe n sat ion Bea ds
There are several different types of thermistor based devices that customers
send in to NIST for measurement. One
of these, which comprise a significant
portion of our workload, is manufactured by Hewlett-Packard/Agilent Technologies. [7] These detectors have an
extra set of thermistor beads that are
used to compensate the HP 432A power
meter reading for thermal drift. Because
of how the compensation beads are situated in the circuit of the detector, there
should not be any interaction between
the compensation beads and the measurement beads. [8] However, it has
recently been discovered that this is not
true. Figure 2 shows the results for measuring an Agilent model 8478B thermistor
detector with and without the compensation beads biased.
0.995
0.990
0.985
system, a detector with a known effective
efficiency (ηe) is connected to the test
port of the power splitter. From the
known ηe of the standard, the reflection
coefficient of the standard, the reflection
coefficient looking back into the splitter,
and the power measured in both the
standard and the monitor detector, a
value can be determined (Ka) for each
measurement frequency which relates
the power available at the test port to the
power measured in the monitor detector.
With Ka known, the DUT can be measured. Mathematically, the calibration is
represented by:
Biased
Not Biased
0.980
0.975
Biased
0.970
Not Biased
0.965
0.960
0.955
10
15
20
0.950
0.945
0
5
10
15
20
Frequency (GHz)
F i g u re 2. Effect of biasing the compensation bead.
MEASURE
|
47
TECHNICAL PAPERS
1.025
1.02
1.025
1.015
1.02
1.01
PM1
1.015
PM2
1.005
1
0.995
0.99
PM3
1.01
PM4
1.005
1
0.995
0.985
0.1
0.2
0.3
0.4
0.99
0.985
0.1
0.5
0.6
0.7
0.8
0.9
1
Frequency (MHz)
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Frequency (MHz)
F i g u re 3. Measurement results for different power meters.
It can be seen in Fig. 2 that a definite
shift occurred in the results when the
compensation beads were biased compared to when they were not biased. This
difference is detector-dependent and can
be as large as 0.015 (in effective efficiency), which is much larger than the
uncertainty of the measurement (approximately 0.006 to 0.0075). This effect is
generally seen in the range from 15 GHz
to 18 GHz. We believe that leakage of
the RF signal between the sets of beads
through some of the blocking capacitors
in the detector is causing the effect. Note
that at frequencies below 15 GHz there
is basically no difference between the
measurements. For detectors that have a
lower operating frequency range, 1 MHz
to 1 GHz, there are no differences
caused by the different biasing conditions.
4 .2 E f f e c t o f D e t e c t o r R F L e a k a g e
on P o w er M eter s and D VM s
Filtering of the RF signals from the dc
leads in the detectors at low frequencies
is difficult. Not all of the RF energy is
absorbed by the thermistor measurement
beads and not all of the unabsorbed
energy is blocked from the dc leads going
to the power meter and on to the DVMs.
This energy creates problems.
The power meters basically do not
react to the extra signal; they may attenuate the signal, but that is about the
48
|
MEASURE
extent of the interaction. The DVMs will
react to the RF leakage signal. The frequency of the signal is important in how
it interacts with the DVM. Figure 3
shows the results for measurements
using several different power meters.
Power Meters 1 and 2 (PM1 and PM2)
are of the same design and PM3 and PM4
are of the same design which is different
than PM1 and PM2.
Several things are happening to the RF
signal in these measurements. The power
meter is filtering the RF signal that has
leaked through the detector, and the
DVMs are reacting to the extraneous
signal. PM1 and PM2 do not filter the
signal as much as PM3 and PM4. It is
also apparent that PM1 and PM2 do not
filter the signal equally. To support our
assertion that it is the RF signal affecting
the DVMs, measurements were made
with a nominal 100 Hz low pass filter in
the leads between the power meter and
the DVMs. In these measurements the
initial results from power meters like
PM1 and PM2 were consistent with the
results from PM3 and PM4. Further
measurements are underway to verify
this effect.
5. Unc er ta inty C om pone nts
There are several main uncertainty components for this system. These include
the contributions from the calibration
standard, the mismatch correction, the
system electronics, device variability,
and a component from an external power
meter, if one is used with a thermoelectric detector. The main components for
the 50 MHz to 18 GHz system have been
detailed in references [2] and [3]. The
expanded uncertainty (coverage factor =
2) runs
PM1 from approximately 0.003 to
0.0076
PM2 (50 MHz to 18 GHz) for a therPM3detector, and from approximately
mistor
0.013PM4to 0.016 for a thermoelectric
sensor.
The differences in uncertainty components for this new system occur in the
100 kHz to 50 MHz range. The main differences are from the calibration standard used below 50 MHz (detailed in the
calibration standards section above) and
from the different VNA that must be
used below 50 MHz for reflection coefficient measurements. This VNA has different residual directivity, source match
and reflection tracking terms. These are
accounted for in the uncertainty of the
reflection coefficient measurements,
which are propagated into the mismatch
uncertainty. [2] For a thermistor sensor,
the expanded uncertainty ranges from
0.003 to 0.0045 and for a thermoelectric
the uncertainty is approximately 0.0126
to 0.013 (100 KHz to 50 MHz). The
uncertainty analysis is still being refined.
6 . M eas ur e m e n t R e s u l t s
We have compared measurements from
the new system to measurements made
by the voltage and impedance technique
(Fig. 4) and to measurements made on
the NIST 0.05 GHz to 50 GHz Direct
Comparison System (Fig. 5).
The results in Fig. 4 show fairly good
agreement between the two different
measurement techniques. The vertical
bars are error bars (for three different
frequencies) based on the uncertainty of
measurements on the new system. It can
be seen that the difference in the two
results is much less than the uncertainty.
We do not yet know what is causing the
systematic difference between the different results. More than likely, it is the
result of slight errors in the values for the
calibration standards.
The results from the new system are
showing very good agreement with the
existing 0.05 GHz to 50 GHz system.
The error bars in Fig. 5 show the uncerwww.ncsli.org
TECHNICAL PAPERS
tainty for the measurements from the
new system at a few selected points.
There are slight deviations at the higher
frequencies, which are in the normal
range of differences for measurements at
these frequencies and are1.006
much less than
the uncertainty of the measurement.
1.006
1.004
1.002
1.004
7. Conclus ion
1.002
A new system for measuring
microwave
1.000
power (effective efficiency) has been
0.998
developed at NIST. This is
the first NIST
system able to cover the entire
frequency
0.996
range of 100 kHz to 18 GHz. Also, this
0.994
is the first NIST system able
to measure
thermoelectric detectors 0.992
below 50 MHz.
Several factors became important in the
0.990
design of the system. These included the
RF leakage through the detectors
and the
0.988
subsequent effects on the power meters
0.986
and DVMs, and the effect of biasing the
0.984
compensation bead in thermistor
detec0.1
tors that have one. The uncertainty of the
system either uses the existing analysis
from the NIST 0.05 to 50 GHz system at
the higher frequencies or includes terms
for the different calibration standard and
reflection coefficient determination for
the lower frequencies. The new uncertainties are comparable to the existing
ones. Finally, measurements on the new
system agree very well with measurements from the existing direct comparison system and from the voltage and
impedance technique.
0.99
0.98
0.97
0.96
0.998
0.996
Voltage/Impedance
New System
0.994
0.992
Voltage/Impedance
0.990
New System
0.988
0.986
0.984
0.1
1
10
100
Frequency (MHz)
F i g u re 4. Comparison of measurements on the new system to the voltage/impedance
1
10
100
technique.
Frequency (MHz)
1
0.99
1
1.000
0.98
0.97
0.96
8 . Re f e re n c e s
0.95
0.94
0
[1] Hewlett-Packard, “Fundamentals of RF
and microwave power measurements,”
Hewlett-Packard
Application
Note 12
64-1,
2
4
6
8
10
August 1977. [Note: This Application
Note has been updated by Agilent Technologies as four separate documents (AN
1449-1/2/3/4) available from the web
site: agilent.com]
[2] M.P. Weidman, “Direct Comparison
Transfer of Microwave Power Sensor
Calibrations,” NIST Technical Note
1379, January 1996.
[3] J.R. Juroshek, “NIST 0.05-50 GHz
Direct-Comparison Power Calibration
System,” Proceedings IEEE CPEM 2000,
May 2000.
[4] J.R. Juroshek, “A Direct Calibration
Method for Measuring Equivalent
Source Mismatch,” Microwave Journal,
pp. 106-118, October 1997.
Old System
New System
0.95
Old System
14
16
18
20
0
2
New System
0.94
4
6
8
10
12
14
16
18
20
Frequency (GHz)
F i g u re 5. Comparison of results from the new 100kHz to 18 GHz system to those from the
old 0.05 GHz to 50 GHz system.
[5] A.Y. Rumfelt and L.B. Elwell, “Radio
frequency power measurements,” Proc.
IEEE, vol. 55, (6), pp. 837-850, June
1967.
[6] F.R. Clague, “Coaxial Reference Standard for Microwave Power,” NIST Tech.
Note 1357, April 1993.
[7] Products or companies named here are
cited only in the interest of complete scientific description, and neither consti-
tute nor imply endorsement by NIST or
by the US government. Other products
may be found to serve just as well.
[8] Technical Manuals for the HP Model
8478B and 478A Thermistor Detectors.
[Available from Agilent Technologies
web site under Technical Support,
Manuals and Guides: home.agilent.com/
agilent/home.jspx?cc=US&lc=eng]
MEASURE
|
49
TECHNICAL PAPERS
Remote Time Calibrations
via the NIST Time Measurement
and Analysis Service
M i c h a e l A . L o m b a rdi a nd A n dre w N. N o vi c k
A b s t r a c t : The National Institute of Standards and Technology (NIST) now offers a new remote calibration service
designed to assist laboratories that maintain an accurate local time standard. The service monitors the local time standard by continuously comparing it to the national time standard and reports the comparison results to the customer
in near real-time. This new service, called the NIST Time Measurement and Analysis Service, or TMAS, works by
making simultaneous common-view measurements at NIST and at the customer’s laboratory with up to eight Global
Positioning System (GPS) satellites. Each customer receives a time measurement system that performs the measurements and sends the results to NIST via the Internet for instant processing. Customers can then view their standard’s
performance with respect to NIST in near real-time, using an ordinary web browser. Time is measured with a combined standard uncertainty of less than 15 nanoseconds, and frequency is measured with an uncertainty of less than
1 #10–13 after 1 day of averaging. This paper describes the multi–channel GPS common–view technique used by the
service and the measurement system sent to each customer. It also explains how NIST calibrates each measurement
system prior to shipment, how measurement results are reported to the customer, and how the measurement uncertainties are estimated.
There is a small but growing demand for
calibration laboratories and research
facilities to maintain a high accuracy
time standard. This requires the laboratory to continuously generate a 1 pulse
Michael A. Lombardi
and
Andrew N. Novick
Time and Frequency Division
and Technology1
50
|
MEASURE
per second (pps) on-time signal, and for
laboratories in the United States, to be
able to state the uncertainty of that signal
with respect to the Coordinated Universal Time (UTC) scale maintained at the
National Institute of Standards and Technology (NIST), known as UTC(NIST).
Once the uncertainty of the 1 pps signal
is known, it can then be used as a standard for traceable measurements of time
interval and/or frequency, or as a synchronization source for other timing
systems. High accuracy 1 pps signals are
normally generated by either a cesium
oscillator or a Global Positioning System
disciplined oscillator (GPSDO). Cesium
oscillators are primary laboratory standards that physically realize the base unit
of time interval (the second) as defined
by the International System (SI). How-
1 This paper is a contribution of the United States government, and is not subject to copyright.
An earlier version of this paper appeared in the 2006 NCSLI Conference proceedings. The identification of commercial equipment is for purposes of illustration only, and does not imply
endorsement by NIST or by NCSL International.
www.ncsli.org
TECHNICAL PAPERS
ever, they still need to be synchronized
before serving as a time standard.
GPSDOs are devices that usually contain
a quartz or rubidium oscillator whose
outputs are continuously steered to agree
with signals from the Global Positioning
System (GPS) satellites. In contrast to a
cesium oscillator, a GPSDO is inherently
on-time, and can produce a 1 pps signal
that is usually well within 1 µs of UTC.
However, because it is not usually possible to measure the time offset of a
GPSDO with respect to UTC(NIST), laboratories are often limited to using and
trusting the number quoted on the manufacturer’s specification sheet as an
uncertainty figure.
Laboratories that want their time standards calibrated against UTC(NIST) to
accuracies better than 1 µs have historically had several options, all of which
have some shortcomings. Customers
sometimes ask to send their cesium
oscillator to NIST for calibration, but
this is normally not a good solution, nor
is it practical. NIST offers several frequency calibration services for cesium
oscillators that are sent to Boulder
(Service IDs 77100C, 77110C, and
77120C) [1], but time information is lost
during the shipment to NIST and the
return shipment to the customer, and the
cesium would need to be resynchronized
when it returns to the customer’s lab. In
fact, when the device returns to the customer, even the frequency of the device
might be substantially different from
what it was during the calibration. A
GPSDO can be sent to NIST for delay
calibrations (Service ID 76120S). [1]
This works well if the antenna and cable
are calibrated along with the receiver.
However, due to local reception conditions, the device might perform differently at the customer’s site than it did at
NIST, and the customer will be without a
time reference during the interval when
the unit is gone from their laboratory.
The NIST services described in the
above paragraph follow the traditional
model, common in most fields of metrology, where the device under test (DUT)
is sent to another laboratory for calibration. In these cases, the DUT is sent to
NIST, where it is calibrated and then
returned to the customer along with a
report containing the measurement
results and an uncertainty statement.
This calibration is typically repeated at
an interval determined by the customer,
for example, once every year. The field of
time and frequency typically uses a different model, based upon remote calibration. Unlike the traditional model, a
remote calibration does not require the
customer to send their DUT to NIST.
Instead, the DUT remains in place at the
customer’s site, and NIST sends a measurement system to the customer. The
measurement system then collects data
that are sent back to NIST for processing, and the calibration can last for as
long as the customer wants it to last.
Laboratories that want their standard to
be continuously monitored by NIST can
do so by subscribing to a remote calibration service and have their standard continuously compared to UTC(NIST) every
day of the year.
NIST has offered remote frequency
and time calibration services since 1983.
[2] The original remote time calibration
service, called the Global Time Service
(GTS), was launched that year and continues to serve a number of customers.
However, its technology is now outdated
in some respects. For example, there are
gaps in the measurement data because
the satellites are not continuously
tracked. Instead, satellite data are
recorded during a series of scheduled
tracks that last for only 13 minutes each,
and the single-channel receivers supplied
to some GTS customers track just one
satellite at a time. Perhaps more importantly, the GTS does not allow customers
a convenient way to view their measurement results until they receive their
monthly reports in the mail. With today’s
technology, it seems the ultimate solution
to a customer’s time measurement
problem would be to have their standard
compared to UTC(NIST), 24 hours a
day, 7 days a week, with the results continuously updated via the Internet so that
they can easily be accessed from anywhere. This is the solution provided by
the new NIST Time Measurement and
Analysis Service (TMAS), the subject of
this paper. The TMAS offers measurement uncertainties that are essentially
equivalent to the GTS, but it costs significantly less, and has the advantage of
making its measurement results available
to customers in near real-time via the
Internet.
2. P h y s i c a l D e s c r i p t i o n o f t h e
T M A S M e a s u re m e n t S y s t e m
The TMAS was announced in late 2005
and assigned a Service ID of 76101S by
the NIST calibration office. [1] The
service shares hardware technology previously developed for the NIST Frequency Measurement and Analysis
Service (FMAS) [3], and software technology previously developed for the
Interamerican Metrology System (SIM)
time and frequency comparison network.
Thus, the same technology delivered to
TMAS customers has been proven by
continuously comparing the national
time scales of the National Research
Council in Canada, UTC(NRC), and the
Centro
Nacional
de
Metrologia
(CENAM) in Mexico, UTC(CNM), to
each other and to UTC(NIST), with
excellent results. [4]
Customers who subscribe to the
TMAS receive a measurement system
consisting of an industrial rack-mount
computer, an LCD monitor, and a keyboard with an integrated trackball
(Fig. 1). A time interval counter with a
single shot resolution of about 30 ps and
an eight-channel GPS receiver are
embedded inside the computer case. [3]
The system is assembled by NIST prior
to shipment and is easy to install. The
customer is required only to connect four
cables to the back panel of the system, as
listed in Table 1. When signals are connected and the unit is powered on, it will
begin taking measurements and sending
data back to NIST.
F i g u re 1. The TMAS measurement
system.
MEASURE
|
51
TECHNICAL PAPERS
3. T h e C o m m o n- Vi e w M e a s u re m e n t Te c h n i q u e
Connector
Input Signal Type
Description
Counter
Time Base
BNC
The time interval counter
requires either a 5 or 10 MHz
sine wave signal as its external
time base. This can often be
obtained from the same DUT
that provides the time
standard. This connection is
made with coaxial cable
(typically RG-58).
Time
Standard
BNC
The customer’s 1 pps time
standard is connected to the
measurement unit using a
coaxial cable (typically RG58). The delay of this cable
must be measured by the
customer and entered into the
system software.
GPS
Antenna
TNC
The GPS antenna and cable
are included with the system
and calibrated at NIST prior
to shipment, and a delay value
is already entered into the
system (Section 4). The length
of the antenna cable is
specified by the customer
before the calibration is
started. After the system
arrives at the customer’s site,
the customer is responsible for
mounting the antenna on a
rooftop location with a clear
view of the sky on all sides.
The antenna is small and easy
to mount.
Network
Ethernet
An Ethernet interface is used
to connect the system to the
Internet. The customer is
required to provide an alwayson Internet connection with a
dedicated IP address. The
system transmits measurement
data using the file transfer
protocol (FTP), and ports 20
and 21 must be left open if the
system resides behind a
firewall.
Ta bl e 1. TMAS input signals.
52
|
MEASURE
The TMAS employs the common-view measurement technique
to compare time standards located at remote locations from
each other. Ideally, a comparison between two time standards
would be made by bringing them into the same laboratory and
connecting them both to some type of phase comparator, usually
a time interval counter. If bringing the time standards together
into the same lab is not practical or desirable, the difference
between the two time standards can still be measured by simultaneously comparing both standards to a common reference
signal that can be received at both sites. Both sites record their
measurements and exchange their results, and the results are
subtracted from each other to obtain the time difference
between the two standards. The common-view signal can be
thought of as a transfer standard and its value drops out of the
final measurement result.
To visualize how the common-view technique works, imagine
two people living at opposite ends of a small town who want to
compare the time displayed by the grandfather clocks in their
living rooms. This would be an easy problem to solve if they
could get the clocks together in the same place and compare
them side by side. However, moving the clocks would be difficult and is not practical or desirable. Therefore, each person
agrees to write down the time displayed by their clock when a
fire whistle (located midway between them) blows in their
town, an event that happens periodically. After writing down the
readings, they call or email each other and exchange the time
readings. If the first clock read 12:01:35 and the second clock
read 12:01:47, then simple subtraction tells them that the
second clock was 12 seconds ahead of the first clock when the
fire whistle blew. The time when the fire whistle blew is unimportant. It only matters that it was heard at the same time, and
that a simultaneous measurement was made at both houses. If
so, the measurement reveals the time difference between the
two grandfather clocks and the comparison was successful. [5]
The common-view technique has been used in the time measurement world for many decades, with a number of different
types of signals used as transfer standards. One notable
common-view measurement involved radio station WWV. From
1955 to 1958, the United States Naval Observatory (USNO) in
Washington, D.C. and the National Physical Laboratory (NPL)
in Teddington, United Kingdom made simultaneous commonview measurements of the signals broadcast from WWV, which
was then located in the Washington, D.C. area. The USNO compared WWV to an astronomical time scale (UT2), and NPL
compared WWV to the new cesium standard they had just
developed. The resulting measurement helped the USNO and
NPL equate the length of the astronomical second to the atomic
second, eventually leading to the atomic second being defined
as the duration of 9,192,631,770 energy transitions of the
cesium atom. [6] In later years, common-view measurements
were made with a variety of signals serving as transfer standards, including LORAN-C and television broadcasts, 60 Hz
power line signals, and even pulses from optical pulsars. [7]
Major advances in accurate common-view measurements
began after the first GPS satellite was launched in 1978. Signals
from the GPS satellites were a nearly ideal common-view referwww.ncsli.org
TECHNICAL PAPERS
ence because there was a clear path
between the transmitter and receiver,
and because the lengths of the two paths
between the transmitter and receivers
were nearly equal. Common-view GPS
measurements began at NIST (then
known as NBS) shortly after the first
GPS satellite was launched [8], and as
previously mentioned, a common-view
service was in place by 1983. [2] The
performance of common-view GPS
measurements was some 20 to 30 times
better than results previously obtained
using LORAN-C as a transfer standard
[9], and the common-view GPS technique soon played a central role in the
international calculation of UTC performed by the International Bureau of
Weights and Measures (BIPM), as it does
to this day. [10]
Common-view GPS comparisons use
one or more GPS satellites as the
common-view reference (Fig. 2). There
are several variations of the technique,
but all have the same objective, to
compare time or frequency standards
located at remote locations. The
common-view method involves a GPS
satellite (S), and two receiving sites (A
and B), each containing a GPS receiver,
a time interval counter, and a local time
standard. The satellite transmits a time
signal that is nearly simultaneously
received at A and B, and a measurement
is made at both A and B that compares
the received GPS signal to the local time
standard. Thus, the measurement at site
A compares the GPS signal received over
the path dSA to the local clock, Clock A –
S. Site B receives GPS over the path dSB
and measures Clock B – S. The two
receivers then exchange and difference
the data. Delays that are common to both
paths dSA and dSB cancel out, but delays
that aren’t common to both paths contribute uncertainty to the measurement.
The result of the measurement is (Clock
A – Clock B) with an error term of (dSA
– dSB). Thus, the basic equation for
common-view GPS measurements is:
(Clock A – S) – (Clock B – S) =
(Clock A – Clock B) + (dSA – dSB) (1)
The components that make up the (dSA
– dSB) error term can be measured or
estimated (Section 8) and applied as a
F i g u re 2. Common-view GPS.
correction to the measurement and/or be
accounted for in the uncertainty analysis.
The (dSA – dSB) error term includes not
only delays from the satellite to the
receiving antennas, but also delays that
take place after the signal is received.
Therefore, a key to a successful measurement is to have well understood and
characterized delays at each site. This
means that the common-view systems
must be calibrated so that their relative
delays are as close to zero as possible.
The calibration of TMAS units is done at
NIST prior to shipment to the customer,
and is discussed in Section 4.
3 .1 C o mm o n- Vi e w an d Tr a c e a b i l i t y
For obvious reasons, the common-view
technique simplifies a laboratory’s task
of establishing traceability to the SI. Calibration laboratories are generally
required to establish traceability of their
own measurement standards and measuring instruments to the SI by means of
an unbroken chain of calibrations or
comparisons. The link back to the SI is
normally achieved through measurements that can be traced to the measure-
ment standards maintained by a national
metrology institute (NMI), the role filled
by NIST in the United States. Therefore,
laboratories can establish traceability to
the SI by sending their standard to NIST
for calibration, or to another laboratory
that has had its standard calibrated by
NIST (which of course introduces
another “link” in the traceability chain).
Even then, however, traceability is established only at a given point in time, and
needs to be periodically reestablished.
[11] For example, if a standard had been
calibrated by NIST ten years ago, a laboratory auditor or assessor would probably not consider that to be sufficient
evidence to establish traceability today.
The TMAS completely solves the
traceability problem. If we equate the
TMAS to the model described in Section
3 above, Clock A is the time standard
maintained at the customer’s site, and
Clock B is the national time standard
maintained by NIST. Thus, the TMAS
makes it possible to continuously establish traceability by making continuous,
direct comparisons against the national
standard. This means that the traceability chain back to the NMI contains only
one link [12], which is the optimal situation for obtaining the best measurement
results.
4. C a l i b r a t i o n o f M e a s u re m e n t
S y s te ms Pr i or t o Sh ip me nt
f ro m N I S T
Each measurement system is calibrated
at the NIST Boulder laboratories prior to
being shipped to the customer. The calibration is done by the common-clock
method, where the system under test and
the reference system at NIST are both
measuring the same clock, a 1 pps signal
from the UTC(NIST) time scale (Fig. 3).
The customer’s system is installed at
NIST using the same antenna and cable
that will be shipped to the customer. The
antenna is attached to a previously surveyed mounting pole whose coordinates
are known to within an uncertainty of
less than 20 cm. The length of the baseline between the customer’s antenna and
the reference antenna at NIST is about
6 m. The calibration lasts for 10 days and
results in an average delay number, DRx,
that is entered into the TMAS system
prior to shipment to the customer.
MEASURE
|
53
TECHNICAL PAPERS
6m
GPS Antenna
GPS Antenna
GPS Receiver
GPS Receiver
1 pps
1 pps
Time Interval
Counter
Start
Time Interval
Counter
Stop
Start
Stop
UTC(NIST)
F i g u re 3. A common-view common-clock calibration of a TMAS measurement system.
The time deviation σx(τ) [13], of the
common-clock calibrations is typically
0.2 ns or less at an averaging period of 1
day. Figure 4 shows results for a calibration, where the peak-to-peak variation of
the 10 minute averages was less than 10
ns, the average delay DRx was equal to
41.1 ns, and σx(τ) was equal to 0.16 ns
at an averaging time of 1 day. There are
some outliers in the data, but there
appears to be no significant slope or
trend. However, the results of a
common-view, common-clock calibra-
tion will vary slightly when repeated multiple times, introducing a systematic
error that must be accounted for in the
uncertainty analysis. This will be discussed further in Section 8.
5. Te c h n i c a l D e t a i l s o f t h e T M A S
S o f t w a re a n d H a r d w a re
The GPS receiver used by the TMAS
simultaneously tracks up to eight GPS
satellites and outputs a 1 pps signal that
is compared to customer’s time standard
with a time interval counter. The receiver
46
Common-view, common-clock delay calibration of TMAS unit
45
44
43
Nanoseconds
42
41
40
also provides data used to produce a time
offset reading for each individual satellite, and these readings are displayed on
the system monitor (Fig. 5). Data are
stored in a file containing a header with
the current system settings, and GPS
data contained in a 32 # 144 matrix.
The 32 columns represent the GPS satellites, with each satellite’s data stored in
the column whose number equals its
pseudo-random noise (PRN) code. The
144 rows represent the number of 10
minute periods in 1 day. At the end of
each 10 minute period, the averaged data
are sent via the file transfer protocol
(FTP) to a NIST web server, where they
are reduced and displayed on-the-fly
(Section 6) when requested by a customer. As many as 11 520 minutes of
data (144 segments # 10 minute tracks
# 8 satellites) can be collected per day,
with no dead time or gaps between measurements. This exceeds the maximum
amount of data collectable by the GTS
with a single-channel receiver by a factor
of about 18.
Note that the software installed on the
customer’s measurement system only
collects data and sends it to NIST; it does
not perform the common-view data
reduction. This is done by web-based
analysis software developed at NIST as a
group of common gateway interface
(CGI) applications written with a combination of a compiled BASIC scripting
language and a Java graphics library. The
software can process up to 200 days of
data (28 800 10-minute segments) and
display them on one graph. It quickly
aligns the common-view tracks where
both NIST and customer viewed the
same satellite at the same time and performs the common-view subtraction for
each aligned track. A time difference,
TD, for a single 10 min track is computed as
8
39
TD =
38
37
53814
53813
53812
53811
53810
53809
53808
53807
53806
53805
53804
36
(SatA i − SatBi )
∑
i =1
,
(2)
CV
where SatAi is the series of individual
satellite tracks recorded at site A, SatBi is
the series of tracks recorded at site B,
and CV is the number of satellite tracks
common to both sites.
Modified Julian Dates (03/10/2006 to 03/19/2006)
F i g u re 4. Results of a 10-day TMAS measurement system calibration.
54
|
MEASURE
www.ncsli.org
TECHNICAL PAPERS
6. R e p o r t i n g R e s u l t s
t o the Cust omer
F i g u re 5. The TMAS measurement system displays the collected GPS readings.
Because all of the data collected by
TMAS customers are uploaded to a
NIST server, customers can request and
view the data whenever they wish.
Requests are normally processed within
a fraction of a second and can be made
using any Java-enabled web browser
from any Internet connection, through a
password protected web site. The data
are graphed as either 1 hour (Fig. 6) or 1
day averages, and the web-based software computes both the time deviation,
σx(τ), and Allan deviation, σy(τ) [13], of
the entire data set. In addition, 10
minute, 1 hour, or 1 day averages can be
copied from the web browser and pasted
into a spreadsheet or other application if
the customer wants to perform further
analysis. At the laboratory’s request,
NIST can also provide signed paper
copies of TMAS reports. These reports
are issued monthly, but contain essentially the same information that is available on-line.
The TMAS is a near real-time
common-view system, which is a tremendous benefit to the customer. During
normal operation, the data will be
updated every 10 minutes, meaning that
customers can view their time difference
with respect to UTC(NIST) within
minutes after the measurement was
made. Near real-time common-view
systems have been implemented previously in Asia [14] and in the SIM region
[4], but they are still the exception rather
than the rule. Some common-view services do not report results to the customer
for days or weeks after the measurements were made.
7. Fi el d Te s t s
F i g u re 6. Viewing TMAS data with a web browser.
Figure 7 shows the results of a six month
comparison (October 2005 to March
2006) between the Sandia National Laboratories' primary time standard and the
UTC(NIST) time scale. The Sandia standard is a cesium oscillator located in
Albuquerque, New Mexico, a distance of
about 561 km from the NIST laboratories in Boulder, Colorado. The red line
shows the actual measurement data, and
the blue line is a linear least squares fit.
The slope of the least squares line is
about 1.7 ns per day. This indicates that
MEASURE
|
55
TECHNICAL PAPERS
850
Sandia Labs Primary Time Standard – UTC(NIST)
Sandia Labs Primary Time Standard – UTC(NIST)
Uncertainty
Component
800
Nanoseconds
750
700
Measurement Data
4
GPS antenna
coordinates error
3
Equipment delay
changes
dueLine
to
Linear
Least Squares
environmental
factors
3
53830
53820
53810
53800
53790
53780
53760
53770
53740
53750
53720
53820
53730
53830
53710
53810
53700
53800
53690
53790
53670
53770
53680
53780
53660
53760
500
Measurement Data
Linear Least Squares Line
53650
53750
53730
53640
53740
53720
53710
F i g u re 7. TMAS comparison (six months) between the time standard at Sandia and
UTC(NIST).
10
Propagation delay
changes due to
multipath
2
Errors in modeled
ionospheric
corrections
2
Error in cable
delay measurements made at
customer’s site
1
Resolution of
instrumentation
0.05
0
Nanoseconds
-10
-20
-30
-40
53794
53784
53779
53774
53769
53764
53759
53754
53749
53744
-60
Ta bl e 2. TMAS estimated Type B uncertainties.
UTC(NRC) – UTC(NIST)
-50
53789
53700
53690
550
53680
Calibration of
TMAS measurement unit
at NIST
650
600
53670
Uncertainty
(nanoseconds)
age area of this estimated uncertainty,
typically within 5 ns, which helps to validate the TMAS performance.
F i g u re 8. Comparison between UTC(NRC) and UTC(NIST).
8. TMA S Unc erta inty Ana ly si s
the Sandia standard has a mean frequency offset of 1.9#10–14 with respect
to UTC(NIST).
As described earlier, the TMAS technology has also been field tested by comparing UTC(NIST) to the time scales of
other NMIs in the SIM region. [4] Figure
8 shows the result of a 41 day comparison
between
UTC(NIST)
and
UTC(NRC), the Canadian national standard, over the 2471 km baseline between
Boulder and Ottawa, Canada. NIST and
NRC each contribute data to the BIPM
that are used to help derive the international UTC time scale. The BIPM publishes these data monthly in their
56
|
MEASURE
Circular-T document. [15] Figure 8
shows the results of the daily comparisons made with the TMAS technology in
blue, and the “official” numbers from the
BIPM Circular-T reported at five-day
intervals in red. The Circular-T values
are obtained with common-view GPS,
but are made by different receivers and
with the benefit of some extensive post
processing, with results reported anywhere from two to eight weeks after the
measurements are made. The blue values
have error bars reflecting the estimated
15 ns uncertainty of the TMAS (analysis
is provided in the next section). The Circular-T values are well within the cover-
Estimating the uncertainty of the TMAS
involves evaluating both the Type A and
Type B uncertainties as described in the
ISO standard. [16] Brief examples are
given here for both time and frequency.
8 .1 An a l ysis of Ti me U n c er t a in t y
To evaluate the Type A time uncertainty,
we use the time deviation σx(τ), at an
averaging time of 1 day. The time deviation is an industry standard statistic [13]
that is calculated automatically by our
web-based software. Using the data displayed in Fig. 7, we obtain a Type A
uncertainty of 1.2 ns between NIST and
Sandia, over a baseline of 561 km. This
www.ncsli.org
TECHNICAL PAPERS
4
10-day common-clock calibrations of a TMAS system
Nanoseconds
3
2
1
0
53830
53820
53810
53800
53790
53780
53770
53760
53750
53740
53730
53720
53710
53700
53690
53680
53670
53660
53650
53640
53630
-1
Modified Julian Dates (September 2005 - March 2006)
F i g u re 9. Results of consecutive 10-day common-clock calibrations made over a 190 day
interval.
uncertainty will increase over longer
baselines, but is typically about 1.5 ns for
the 2471 km baseline between NIST and
NRC. As a result, we expect the Type A
time uncertainty to be less than 2 ns for
all TMAS customers in the continental
United States.
The Type B evaluation is more difficult, but we have identified seven components that can potentially introduce
systematic errors that are summarized in
Table 2 and discussed in more detail in
Sections 8.1.1 through 8.1.7. Some Type
B uncertainties can also get larger as a
function of the length of the baseline, but
the estimates provided here should be
applicable for all TMAS customers in the
continental United States, where the
baseline length should not exceed 3000
km. Due to the nature of common-view
measurements, any systematic error that
is common to both sites will cancel out,
so all the Type B components listed here
relate to uncertainties that can affect one
site differently from the other. All type B
uncertainties are treated as normal distributions. In the case of antenna coordinates, we assume that the customer will
be able to survey their antenna’s position
to within an uncertainty of 1 m. If this is
not true, the combined time uncertainty
of the TMAS will increase, as explained
in Section 8.1.2.
8 .1 .1 C a l i b r a t i o n o f T M A S
M e a s u r em e n t U n it a t NI S T
As described in Section 4, the 10-day
common-clock calibrations of TMAS
units are typically stable to 0.2 ns or less,
but the results are not necessarily repeatable at different times of the year. For
example, if a common-clock calibration
were continuously repeated, the resulting
estimate of DRx would vary by at least
several nanoseconds, depending upon
which 10-day segment was chosen. [17]
This is illustrated in Fig. 9, which shows
the results of a unit that was continuously calibrated at NIST over a 190-day
interval spanning from September 2005
to March 2006, producing 181 overlapping 10-day segments. During this interval, the peak-to-peak variation is nearly 4
ns, and a unit could be shipped with a
DRx value from anywhere within this
range. Based on data collected from
repeated calibrations of several units, we
assign a Type B uncertainty of 4 ns to our
delay calibrations.
8. 1 .2 G P S A n t e n na C o o rd i n a t e s E r ro r
The customer is required to obtain coordinates for the GPS antenna prior to
starting the TMAS measurements. If the
customer has a way to independently
survey the antenna, the resulting coordinates can be typed in to the TMAS software. If not, the TMAS system can
survey the antenna position by averaging
position fixes for 24 hours, a method
that does an excellent job of determining
the antenna’s horizontal position (latitude and longitude) to within less than 1
m. However, GPS does a comparatively
poor job of surveying vertical position
(elevation), and the vertical position
error is usually at least several times
larger than the horizontal position error.
This is because GPS provides earth-centered coordinates and measures the distance between the center of the earth and
the satellite. Vertical position is obtained
with the radius of a model of the earth’s
surface. There is nearly always some bias
in the estimated vertical position due to
local terrain that differs from the model.
We assign a Type B uncertainty of 3 ns
to the GPS antenna coordinates, which
assumes that the customer survey is
within 1 m (the approximate distance
that light travels in 3 ns). However, if the
TMAS self survey is used, this uncertainly will probably be larger, as large as
3 ns per meter for some satellites, but
closer to 2 ns per meter of position error,
on average. Figure 10 shows the result of
20 TMAS antenna surveys conducted at
NIST in Boulder, Colorado, each lasting
for 24 hours. Each survey was done with
the same receiver and an antenna that
had been independently surveyed to an
estimated uncertainty of less than 0.2 m.
The blue line in the figure shows the total
position error in the X, Y, Z coordinates
based on the distance from the known
coordinates, and the red line shows the
error in the vertical position for each of
the 20 surveys.
As shown in Fig. 10, the average position error was 5.37 m, with nearly all of
this error due to error in the vertical
position, which was 5.30 m. The estimated vertical positions were biased
about 4 to 6 m above the actual elevation, resulting in a Type B uncertainty
due to antenna coordinates error that
would typically exceed 10 ns, much
larger than our 3 ns allowance. This
might be an acceptable uncertainty for
many customers, but for the best results,
TMAS customers should have their
antenna elevation independently surveyed to within an uncertainty of 1 m.
MEASURE
|
57
6.5
Position Error (meters)
6
5.5
5
Total Position Error
or in multiple 24-hour surveys of
4.5known GPS antenna coordinates
5
6
7
8
9
10
11
12
4
13
14
15
16
17
18
19
Vertical Position Error
20
21
Antenna surveys (one per day)Position error in multiple 24-hour surveys of known GPS antenna coordinates
3.5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Antenna surveys (one per day)
F i g u r e 10. Position errors (with respect to known coordinates) from 20 TMAS antenna surveys.
8 .1 .3 T M A S E q u i p m e n t D e l a y
C h a n g es D u e t o
E n v i ro n m e n t a l F a c t o r s
GPS receiver, antenna, and antenna
cable delays can change over the course
of time due to temperature and other
environmental factors. The GPS receiver
delay has the largest sensitivity to temperature, but its performance will be
very stable if the laboratory temperature
is well controlled. The receiver temperature is typically just a few degrees Celsius
higher than the laboratory temperature,
with a similar range. However, a sudden
change in laboratory temperature can
sometimes cause the receiver delay to
change by several nanoseconds, usually
returning to its previous delay when the
temperature returns to normal. Smaller
receiver delay changes can occur slowly
over time for reasons that are not completely understood. These delay changes
might be caused by fluctuations in power
supply voltages, vibration, or humidity.
As a result, we assign a Type B uncertainty of 3 ns to account for receiver/
antenna delay changes due to the environment.
The GPS antenna and part of the cable
are outdoors, and are thus subjected to
large annual variations in temperature
(the annual temperature range can
exceed 60 °C in Boulder, Colorado).
Even with this large of a range, the actual
changes in the electrical delay of the
cable due to temperature are insignifi58
|
MEASURE
cant, but can potentially cause the
receiver tracking point to change, introducing phase steps in the data. [18] The
TMAS guards against this possibility by
using a high quality antenna cable with a
low temperature coefficient.
8 .1 .4 P r o p a g a t i o n D e l a y C h a n g e s
Due to Multipath
Errors due to multipath are caused by
GPS signals being reflected from surfaces near the antenna. These reflected
signals can then either interfere with, or
be mistaken for, the signals that follow a
straight line path from the satellite.
TMAS customers are instructed to
mount their antennas in an area with a
clear, unobstructed view of the sky on all
sides, and an antenna is used that was
designed to reject multipath signals. For
these reasons, the uncertainty due to
multipath is usually very small. However,
because some errors due to multipath are
difficult to detect and avoid, we assign a
Type B uncertainty of 2 ns. [19]
8 .1 .5 P r o p a g a t i o n D e l a y C h a n g e s
D ue t o I on os pher i c Co ndi ti o ns
The GPS signals are line of sight, and the
path delay between the satellites and the
receiver can be accurately estimated
from the distance and the speed of light.
However, the signals are bent slightly as
they pass through the ionosphere and
troposphere, changing their propagation
delay. The delay changes are largest for
satellites at low elevation angles. The
GPS satellites broadcast a modeled
ionospheric delay correction that is automatically applied by the TMAS to the
measurements made at both sites.
However, ionospheric conditions are not
identical at both sites (particularly when
it is dark at one site and daylight at the
other), and some common-view GPS
systems apply ionospheric corrections as
measured at each site, instead of using
the broadcast corrections. [19] This
delays the processing of the measurement
by at least one day, but
Total
Positionresults
Error
Vertical
Position Error
reduces
the measurement uncertainty.
Because the TMAS uses modeled ionospheric corrections as opposed to measured corrections, we assign a Type B
uncertainty of 2 ns for ionospheric delay
that should cover all customers in the
continental United States.
8 .1 .6 C a b l e D e l a y M e a s u r e m e n t s
M a d e a t C u s t o m e r ’s L o c a t i o n
When the TMAS unit is installed, the
customer is responsible for measuring
the reference delay, or DREF, and entering this value into the system software.
The reference delay represents the delay
from the local time standard to the end
of the cable that connects to the TMAS
system. This is typically a one-time measurement made by the customer with a
time interval counter. The Type B uncertainty will normally not exceed 1 ns if
proper measurement techniques are
followed.
8 .1 .7 R e s o l u t i o n U n c e r t a i n t y o f
S o f t w are an d I ns tr u me nt at i on
The TMAS software limits the resolution
of the entered delay values to 0.1 ns, contributing an insignificant resolution
uncertainty of 0.05 ns.
8 . 1 . 8 C o m b i n ed T i m e U n ce r t ai n ty
The combined Type B uncertainty, Ub, is
obtained by taking the square root of the
sum of the squares of the estimated
uncertainties listed in Table 2, and equals
6.6 ns. The combined expanded uncertainty Uc is obtained by this equation:
Uc = k Ua2 + Ub2 .
(3)
If we use a coverage factor of k = 2 and
Ua = 2 ns (as discussed in Section 8.1),
www.ncsli.org
TECHNICAL PAPERS
then Uc is equal to 13.7 ns. This figure
has been rounded up to a conservative
service specification of 15 ns that should
be achievable with all customers. In the
case of the 2471 km baseline between
NIST and NRC, these results have been
validated with independent measurements published by the BIPM [15] that
fall well within the TMAS coverage area
(Fig. 8).
8 .2 An alysis of Fre q u e n c y U n c e r t a i n t y
Frequency uncertainty can be estimated
by fitting a least squares linear line to the
data to obtain a mean frequency offset,
Y, and then using 2σy(τ) [13] as the Type
A uncertainty Ua (k = 2 coverage). Since
there is no significant Type B component
for frequency, the combined uncertainty
Uc can be considered as the Type A
uncertainty. The upper and lower bounds
of the coverage area are represented by Y
+ Uc and Y – Uc, respectively. For the 6month data run shown in Fig. 7, the
mean frequency offset is 1.9 # 10–14,
with a k = 2 uncertainty of approximately
1.3 # 10–14 after one month of averaging. The lower and upper bounds of the
coverage area over a one month interval
would be 0.6 #10–14 and 3.2 #10–14,
respectively, with respect to UTC(NIST).
Note that the frequency uncertainty
decreases as the averaging time
increases. The estimated uncertainty
after 1 day of averaging is near 5 #10–14.
9. Summ ary
The NIST Time and Measurement and
Analysis service makes the measurement techniques used for international
comparisons between the world’s best
timing laboratories available to any calibration lab or research facility. The
TMAS offers a combined standard
uncertainty (k = 2 coverage factor) of
less than 15 nanoseconds for time, and
less than 1 #10–13 for frequency after 1
day of averaging. The service is available
though NIST as service number 76101S
at a cost of $750 per month, with a onetime startup fee of $1500. [1]
10. A c kno wl e dg em e nts
The authors thank Bob Graham of
Sandia National Laboratories in Albuquerque, NM, for his assistance in beta
testing a TMAS system, and for the use
of the data shown in Fig. 7. We also
thank the National Research Council in
Ottawa, Canada for the use of the SIM
data shown in Fig. 8.
11 . R e f e re n c e s
[1] “NIST Calibration Services User Guide,”
NIST Special Publication 250, (current
copy is available on-line at http://ts.nist.gov
/ts/htdocs/230/233/calibrations/).
[2] S.R. Stein, G. Kamas, and D.W. Allan,
“New time and frequency services at the
National Bureau of Standards,” Proceedings of the 15th Annual Precise
Time and Time Interval (PTTI) Applications and Planning Meeting, pp. 17-27,
December 1983.
[3] M.A. Lombardi, “Remote frequency calibrations: The NIST frequency measurement and analysis service,” NIST Special
Publication 250-29, 90 pages, June 2004.
[4] M.A. Lombardi, A.N. Novick, J.M.
Lopez, J.S. Boulanger, and R. Pelletier,
“The Interamerican Metrology System
(SIM) Common-View GPS Comparison
Network,” Proceedings of the 2005 IEEE
Frequency Control Symposium, pp. 691698, August 2005.
[5] M.A. Lombardi, A.N. Novick, and R.M.
Graham, “Remote Calibration of a GPS
Timing Receiver to UTC(NIST) via the
Internet,” Proceedings of the 2003 Measurement Science Conference, January
2003.
[6] T. Jones, Splitting the Second: The Story
of Atomic Time, Institute of Physics Publishing, Bristol, UK, pp. 123-126, 2000.
[7] D.W. Allan, H.E. Machlan, and J. Marshall, “Time Transfer using Nearly Simultaneous Reception Times of a Common
Transmission,” Proceedings of the 1972
Frequency Control Symposium, pp. 309316, June 1972.
[8] D.W. Allan and M.A. Weiss, “Accurate
Time and Frequency Transfer During
Common-View of a GPS Satellite,” Proceedings of the 1980 Frequency Control
Symposium, pp. 334-346, May 1980.
[9] D.W. Allan, D.D. Davis, M.A. Weiss,
A.J. Clements, B. Guinot, M. Granveaud,
K. Dorenwendt, B. Fischer, P. Hetzel, S.
Aoki, M.-K. Fujimoto, L. Charron, and
N. Ashby, “Accuracy of International
Time and Frequency Comparisons via
Global Positioning System Satellites in
Common-View,” IEEE Transactions on
Instrumentation and Measurement, vol.
IM-34, no. 2, pp. 118-125, June 1985.
[10] W. Lewandowski, J. Azoubib, and W.
Klepczynski, “GPS: Primary Tool for
Time Transfer,” Proceedings of the IEEE,
vol. 87, no. 1, pp. 163-172, January
1999.
[11] C.D. Ehrlich and S.D. Rasberry, “Metrological Timelines in Traceability,” J. of
Res. of NIST, vol. 103, no. 1, pp. 93-105,
January-February 1998.
[12] M.A. Lombardi and A.N. Novick, “Comparison of the one-way and commonview GPS measurement techniques using
a known frequency offset,” Proceedings
of the 34th Annual Precise Time and
Time Interval (PTTI) Systems and Applications Meeting, pp. 39-51, December
2002.
[13] “IEEE Standard Definitions of Physical
Quantities for Fundamental Frequency
and Time Metrology – Random Instabilities,” IEEE Standard 1139-1999, prepared by IEEE Standards Coordinating
Committee 27 on Time and Frequency,
March 1999.
[14] X. Yang, Y. Hu, Z. Li, X. Li, and X.
Zheng, “An algorithm for a near realtime data processing of GPS commonview observations,” Chinese Astronomy
and Astrophysics, vol. 27, no. 4, pp. 470480, October-December 2003.
[15] The BIPM “Circular-T” reports are
archived at: www.bipm.org
[16] “ISO Guide to the Expression of Uncertainty in Measurement,” prepared by
ISO Technical Advisory Group 4,
Working Group 3, October 1993.
[17] J. Levine, “Averaging satellite timing
data for national and international time
coordination,” Proceedings of the 36th
Annual Precise Time and Time Interval
(PTTI) Systems and Applications
Meeting, pp. 41-52, December 2004.
[18] M.A. Weiss, “Long Term Effects of
Antenna Cables on GPS Timing
Receivers,” Proceedings of the 2000
IEEE Frequency Control Symposium, pp.
637-641, June 2000.
[19] W. Lewandowski and C. Thomas, “GPS
Time Transfer,” Proceedings of the IEEE,
vol. 79, no. 4, pp. 991-1000, 1991.
MEASURE
|
59
REVIEW PAPERS
Gravimetric Calibration
of Volumetric Standards with
L .F. E a so n
A b s t r a c t : Recently, the demand for volumetric measurements with greater accuracy and smaller measurement uncer-
tainties has increased dramatically. In response, the metrology laboratories of the Arizona, Maine, Michigan, and North
Carolina weights and measures programs have established gravimetric calibration capabilities for volume standards
(provers) with capacities up to 100 gallons (500 liters). This collaborative effort with the National Institute of Standards and Technology (NIST), Weights and Measures Division (WMD), has improved volumetric prover calibration
accuracy and uncertainty significantly. Accuracy is a measure of how close a measurement is to the actual value. Uncertainty is a measure of how well the value is known. Smaller uncertainties in laboratory standards have lead to similar
improvements in field prover calibrations up to 2000 gallons. Prover calibration improvements facilitate better meter
calibrations. Consequently, petrochemical terminals can have more confidence in inventory records, reducing inexplicable product losses. Apparent losses that once were ignored because they were less than the existing measurement
uncertainty can now be investigated. Meters can be adjusted before loss totals increase.
Gravimetric volume calibration makes use of existing mass comparator balance technology and mass standards
commonly available in the State metrology laboratories. Any volume up to the limit of available mass standards and
balance capacity can be calibrated without the expense of the multiple volume standard provers typically required for
calibration of non-standard volumes. Thus, metrology laboratories benefit from significant cost savings in addition to
improving their calibration process. Since gravimetric calibrations are traceable to mass, fewer laboratory volume standards must be calibrated by NIST, providing another cost savings to laboratories. Prior to large volume gravimetric
calibration, volume transfer calibration uncertainties (k = 2) from NIST were
reported at 130 ppm (3.1 cubic inches for a 100 gallon prover). Using this
L. F. Eason
uncertainty as a starting point, laboratories then had to add their own process
North Carolina Department
uncertainty factors and reported significantly higher uncertainties for field
of Agriculture & Consumer Services
provers. Using gravimetric calibration, a technician proficient in mass metrolStandards Laboratory
ogy can expect to attain expanded uncertainties (k = 2) of 70 ppm (1.6 cubic
1051 Mail Service Center
inches for a 100 gallon prover). These advances in gravimetric calibration
Raleigh, NC 27699-1051 USA
E-mail: [email protected]
improve volume measurements at all levels, from the laboratory, to the terminal, and to the retail market.
60
|
MEASURE
www.ncsli.org
REVIEW PAPERS
1. The Mark et
Several market-based factors are driving
the need for more accurate volume standard calibrations. With record increases
in the price of petrochemical products,
the potential cost of inaccurate measurements is greater than ever before. The
cost of downtime needed for high
volume petrochemical meter calibrations
also continues to increase. Meter manufacturers have developed new metering
systems with better temperature measurement and compensation systems to
address the need for increased measurement accuracy. More efficient volume
standards, such as dynamic small volume
provers (SVPs) have been developed to
decrease down time for meter testing
(typically known as meter proving within
the industry). These advances in measurement technologies have necessitated
commensurate improvements in volume
standard calibration technology.
According to the March 2005 Petroleum Marketing Monthly, almost 712
million gallons1 of petroleum products
were sold in the United States each day
during the month of December 2004. [1]
The average before tax U.S. refiner price
for this product during the same time
was $1.21(US) per gallon. Even at this
price, petroleum sales exceed $861
million (US) per day. If December sales
were typical of other months, that
equates to over $314 billion (US) per
year. This figure ignores tax revenues and
retail profit margins. The price continues
to rise. Obviously, measurement of this
product is important so that measurement uncertainties should be minimized.
2 . M e te r To l e r a n c e s
In the United States, commercial meter
tolerances are set by the National Conference on Weights and Measures
(NCWM). These consensus standards
are published in NIST Handbook 44.
Petroleum meters are covered as class
0.3 devices under the Liquid Measuring
Device (LMD) code. [2] Lowering these
tolerances for liquid measuring devices
has been proposed as one way to mini1 The SI unit for volume is cubic meters, but
for all practical realizations in the United
States, gallons are used.
6
5
7
1.
Drain valve
2.
Drain slope 5º
3.
Levels
4.
Level cover
5.
Gauge mounting
6.
Rolled bean
7.
Top cone pitch 25º
8.
Reinforcing bands
9.
Bottom core pitch 20º
4
3
8
2
1
9
10.
Adjustable legs
10
F i g u re 1. NIST Handbook 105-3 Field Standard Prover. [3]
mize uncertainties. Several issues need to
be considered before this course of
action is taken.
It is debatable if lowering the tolerances will have any effect on the market.
In the author’s experience, few (if any)
petroleum terminals allow meters to be
in error by the NIST Handbook 44 tolerance of ±0.3 % for routine calibrations
and ±0.2 % for new installations or
repaired meters. It is common for terminals to require meter adjustments for
errors that exceed either ±0.05 or
±0.03 %. Meter technology appears to
allow these adjustments and meters
repeat within this reduced allowance.
However, what about the meter’s accuracy and the accuracy of the standards
used to test these meters?
Like any other measuring device,
petroleum meters can only be as accurate
as the standards that are used to calibrate
them. Historically, these standards have
been either mild-steel or stainless steel
vessels. Test measures ranging from one
to ten gallons are used to calibrate small
capacity meters such as service station
gas pumps. Provers ranging from 25
gallons to 2000 gallons are used to calibrate (prove) meters ranging from high
capacity diesel pumps at truck stops, to
large meters at fuel terminals. Very large
capacity loop provers are used to calibrate the very high flow rate meters used
at refineries and on petroleum pipelines.
This paper will focus on the calibration
of 25 to 2000 gallon volumetric provers
used in the calibration of the midrange
meters.
Calibration of newer technology
devices, known as dynamic small volume
provers (SVPs) or captive displacement
provers (CDPs), will also be discussed.
For the remainder of this paper, these
standards are referred to simply as
provers and SVPs.
3 . Pr o v e r D e s i g n
Provers typically are designed to meet
National Institute of Standards and Technology (NIST) [3] or the American
Petroleum Institute (API) design requirements. [4] The NIST and API design
requirements are very similar (see
Fig. 1). Provers are made of either
carbon or series 300 stainless steel, both
of which have well defined thermal
expansion properties.
Historically, provers have been calibrated by a volume transfer calibration
MEASURE
|
61
REVIEW PAPERS
procedure. Water is transferred into the
prover being tested from other NIST
traceable, calibrated provers. API has
also recognized water draw calibration,
which is the reverse of the volume transfer. Instead of water being delivered from
standards into the unit under test, the
unit under test is filled to the nominal
zero mark and drained into a calibrated
prover. Water draw calibration has often
been used as a field calibration technique
for provers that are too large to transport
to a laboratory for calibration.
For volume transfer and water draw,
the difference between the known
volumes of the standards and the
observed volume of the prover being calibrated is used to either calculate a calibration factor or adjust the prover under
test as close to nominal as possible. Temperature corrections based on the temperature of the water and the coefficients
of expansion for both provers are calculated. The calibrated volume is referenced at either 15 ºC or 60 ºF, depending
on where the prover will be used.
Both NIST and API procedures require
provers to be large enough to hold the
amount of liquid delivered from a meter
at full flow for one minute. Since start up
and stopping flow rates are reduced to
prevent excessive vaporization, over
flow, and foaming of the product, a
prover actually has to be somewhat
larger than the liters or gallons per
minute delivered through the meter. For
example, to deliver one minute at full
flow, a meter delivering 400 gallons per
minute of fuel oil, might need to deliver
an additional 100 gallons at a slower
flow to prevent excessive foaming at the
start and end of the calibration. Therefore, using a 500-gallon prover would be
marginal to test this meter even though it
is larger than the volume delivered in a
minute of full flow from the meter. The
same prover may be acceptable for a
meter of the same capacity delivering
gasoline since the start up and slow
down volumes are less for gasoline.
Provers are typically calibrated “to
deliver” a specified volume. This is necessary so that the prover doesn’t have to
be completely dried between runs.
Instead, a drain time is specified. The
drain time is typically 30 seconds after
cessation of main flow. This wet down
62
|
MEASURE
procedure is done at the time of calibration and every time the prover is used in
the field. After the drain time, the drain
valve is closed and the remaining
product clinging to the walls of the vessel
is allowed to remain in the prover. Thus,
an assumption is made that there is a
repeatable amount of fluid left in the
prover at the beginning of each run. This
necessitates a wet down run that is disregarded each time the prover is used.
Establishing an accurate drain time is
problematic for water draw procedures
when multiple standards are required
since the prover has to be stopped each
time a standard is filled and the next
standard is filled.
4. Sm al l Vo l u m e P ro v e r D e s i g n
There are several problems with the use
of traditional provers at terminal
loading racks. These include:
• In order to hold a full minute flow for
high capacity meters, provers must be
very large.
• Provers are difficult to transport from
one location to the next or to the laboratory for calibration.
• Since provers are large, they require
the lane of meters being calibrated to
be shut down for extended lengths of
time.
• Meter testing at multiple flow rates is
very limited due to the time requirements to fill the large volume.
• Even during a “full flow test,” some
percentage of the volume is actually
pumped at a slower rate for start-up
and slow-down. Thus, the meter
factors calibrated for full flow are
always slightly skewed away from the
actual full flow factor. This can be
significant for kerosene and fuel oil
distillates, since the volume metered
at slow flow may be up to 25 % of
the total volume.
• Since the percentage of slow flow to
full flow will vary between provers of
different volumes, different meter
factors will be calculated, depending
on the nominal volume of the provers
being used.
• There is a possibility of spills from
overflow when a meter is significantly
out of calibration.
• Since a traditional prover is not a
completely closed system, vapor loss
may be a consideration.
• Drain times often differ considerably
depending on the product and the
pump back system. Differences in
drain times affect the wet down condition of a prover and the calibration
by an unknown amount.
To minimize these problems, the
dynamic small volume prover was developed. [5] Rather than a tank of known
volume, the SVP is a displacement type
prover. A piston travels with the flow of
liquid through a cylinder (see Fig. 2).
The calibrated volume of the cylinder is
defined by two optical switches that read
a flag traveling with the piston. The
meter test begins as the flag on the piston
rod passes the first optical switch and
stops as it passes the second calibrated
switch. At the end of the calibrated
length, a poppet valve opens to let
product flow by with minimal obstruction. The piston is then pulled back to a
staging position before the first optical
switch to start the process over again.
Between these two switches, the computer on the SVP counts the pulses generated by the meter. This counter uses
two timers that are slightly offset. At the
end of the calibration run, the pulse
count is interpolated between the two
timers using a technology called double
chronometry pulse interpolation. [6]
This allows a much more precise count
of pulses than the traditional pipe or ball
prover technology. This procedure is
repeated a number of times (typically
five) at each flow rate. The interpolated
pulse values are compared to verify the
repeatability of the meter. If all repeat
within a specified percentage (typically
±0.02 % of volume), the run is
approved, the interpolated pulse counts
are averaged, and a meter factor is calibrated for that specific flow rate.
The meter flow rate capacity versus
the SVPs calibrated volume is very large
compared to the traditional graduated
neck prover. Depending on the manufacturer and model, SVPs may be rated for
flow rates up to 100 times its calibrated
volume. For example, a SVP with a
fifteen-gallon calibrated volume is rated
at 1497 gallons per minute. It would take
a 2000 gallon graduated neck prover to
calibrate a meter with the same flow rate.
Though the SVP tends to minimize
www.ncsli.org
REVIEW PAPERS
Seal Monitor Sensor
On-Board Microprocessor
Optical Detector
Switches
Measuring Section
Invar Rod
Piston Guide
Displacer Actuator
Launch Spring
Main Displacer
Power
Unit
Differential Pressure
Switch for Monitoring
Bypass Seal Integrity
Main Displacer Seals
Guide Tracks
Guide Rods
Outlet Flange
Bypass Valve
Balance Valve
Vertical Lift
Stop
Inlet Flange
Bypass
Actuator
F i g u re 2. Small volume prover from NIST Handbook 105-7. [3]
most of the problems identified with traditional graduated neck provers, there
are issues unique to this technology. With
this high flow rate to calibrated volume
ratio, it is obvious that the calibration of
the SVP is very critical. A small uncertainty in a graduated neck prover calibration has much less affect than the same
uncertainty on a SVP calibration. This is
one of the primary reasons that the
metrology laboratories of several State
Weights and Measures programs are
moving to establish improved gravimetric measuring capabilities for SVPs and
larger volume standards. Other factors
that affect comparisons between meter
calibrations in the field using SVPs and
graduated neck provers are beyond the
scope of this paper.
5. Tr a d i t i o n a l C a l i b r a t i o n
P ro c e d u r e s
As with the calibration of meters, the calibration of provers is limited by the
quality of the calibration standards used.
Traceability to the national standard is
the ultimate goal. Since volume is
derived from mass, in the United States
this chain of traceability must eventually
extend back to the national prototype
kilogram 20 (K 20) at NIST. This platinum iridium mass standard is used to
calibrate multiple levels of mass working
standards. These are then used to propagate the unit of mass to both larger and
smaller working standards by using
complex weighing designs. At NIST,
some of these working standards are
then used to gravimetrically calibrate
small volume standards up to the fivegallon level. (Gravimetric calibration is
covered in detail later in this paper.) Historically, provers larger than 20 liter or
five gallon were calibrated by volume
transfer, using multiple drops from a well
characterized, gravimetrically calibrated
five-gallon slicker plate standard. The
weighing equipment available at the
time, a large equal arm balance, imposed
this limitation. A 100-gallon NIST
working standard was also calibrated by
volume transfer and used to calibrate
100-gallon and larger customer standards. With this procedure, the reported
NIST volume calibration uncertainty
(k = 3) was ±0.02 % of the volume. This
resulted in an uncertainty of ±100 milliliters per 500 liters or ±4.62 cubic inches
per 100 gallons.
Since each link in a calibration chain
adds to the measurement uncertainty,
this ±0.02 % was just the starting point
for additional calibrations. In the United
States, NIST performs relatively few
large volume calibrations compared to
the next level in the national measurement chain. The majority of large volume
calibrations are delegated to private calibration laboratories and the State
metrology laboratories. According to the
2003 NCWM State Laboratory Program
Survey, in 2002, the State metrology laboratories calibrated 6966 test measures,
1053 provers, and 555 glassware standards. The state metrology laboratories
are coordinated and evaluated by the
NIST Weights and Measures Division
(WMD). In addition to NIST WMD
recognition, many of the state metrology
laboratories are now NIST National Voluntary Laboratory Accreditation Program
MEASURE
|
63
REVIEW PAPERS
(NVLAP) accredited. Both of these programs are based on ISO/IEC 17025
requirements.
Each additional step in the traceability
chain between NIST and a prover being
calibrated adds an additional level of
uncertainty to the measurement. Thus,
volumetric transfer calibrations from
these labs started with a ±0.02 % uncertainty for the standards calibrated at
NIST. Each laboratory added a level of
uncertainty based on their process uncertainty to the calibration of a customer’s
prover. NIST WMD has circulated a 100gallon stainless steel prover for interlaboratory comparisons among State
metrology laboratories since 1988. The
average calibration uncertainty reported
by states between 1988 and 1993 was
±7.0 cubic inches or approximately
±0.03 % of the volume. NIST Handbook
44 [2] and virtually all quality programs
require the total standard and calibration
process uncertainty to be less than one
third of the tolerance of the device under
test. Starting with a standard uncertainty
of ±0.03 %, even if the process uncertainty of the meter calibration were negligible, the provers would not be adequate
for meter test tolerances below ±0.1 %.
This is double the tolerance that the terminal loss control engineers use.
Many additional factors affect meter calibration in a field calibration. These include:
• Temperature measurement
• Scale plate readability
• Steel coefficient of expansion
uncertainty
• Evaporation
• Vaporization
• Pressure surges from other trucks
filling
• Differences in pump off time from
calibration drain time
• Prover drain characteristics
• Operator training
• Meter reading sensitivity
• Prover leveling
• Various other human errors
When these factors are all combined, it
is easy to identify another ±0.1 % uncertainty in the meter calibration process
itself. At this level of uncertainty, it was
difficult to achieve the NIST Handbook
44 tolerances, much less the ±0.05 %
uncertainty required by the terminals.
Improvements were needed.
64
|
MEASURE
6. I m pro v e m e n t s a t N I S T
Recognizing the need for improvement,
API worked with NIST in the early
1990’s to improve the measurement
capabilities of the NIST volume laboratory. In addition to facility improvements, a new 600-kilogram capacity
mass comparator was purchased to allow
gravimetric calibration up to 500 liters
or 100 gallons. In a gravimetric calibration, the water in a vessel is weighed. The
volume is then calculated based on the
density of the water with suitable corrections for temperature and air buoyancy.
Thus, an entire step in the traceability
measurement chain is eliminated. A customer’s 100-gallon prover could be calibrated directly from mass working
standards. In addition, the NIST 100gallon working standard prover was calibrated gravimetrically and housed in a
climate controlled laboratory. When this
system was brought on line in 1994, the
NIST volume calibration uncertainty
dropped from ±0.02 % to ±0.004 %.
Volume transfer calibration uncertainties
from the 100-gallon working standard
also improved both from the lower starting uncertainty of the 100-gallon
working standard and the improved
environment used for the calibrations.
This was a very large step forward for
volume measurement in the United
States.
Another benefit realized from this
improvement was that the technology
spread to the State metrology laboratories, including the North Carolina (NC)
Standards Laboratory. The large volume
gravimetric calibration procedure uses
equipment, standards, and procedures
that were already in many of the state
metrology laboratories. According to the
2003 NCWM State Laboratory Program
Survey, in 2002 nearly 80 % of the
375,000 standards tested by the state
metrology laboratories each year are
mass standards. With this workload, the
state laboratories have focused on maintaining their mass calibration capabilities. To meet NIST WMD and NVLAP
mass calibration requirements, these laboratories have well controlled laboratory
environments, well-characterized stateof-the-art mass comparators, very precise
environmental parameter measurement
instrumentation, and well trained
metrologists. As gravimetric calibration
techniques were advanced at NIST, the
state laboratories were in a good position
to adopt the same procedures. The NC
Standards Laboratory had been using
gravimetric calibration for all volumes
up to 20 liters and five gallons with
excellent results. It was a relatively easy
step to increase the range to 100 gallons.
7. Gr a vi m et ri c Ca l ib ra ti on
The original National Bureau of Standards (NBS, changed to NIST in 1988)
gravimetric procedures for large volume
standards were written for equal arm
balances. These procedures were very
cumbersome, requiring transposition or
double substitution weighing of the
volume standard empty, filled, and
drained. Each measurement required
different mass standards and was very
time consuming. With the filled vessel
supported on the hanging pan of the
balance, there were many opportunities
for spills and damage to standards. A
source of distilled or deionized water had
to be near the balance. If water dripped
on the outside of the standard, it had to
be carefully dried off to not affect the
measurement. Due to the time it took,
evaporation was a significant problem.
Gravimetric calibrations above one liter
were uncommon.
The advent of the single pan electronic
balance improved gravimetric calibration
possibilities. As precision electronic
mass comparators increased in capacity
and resolution, gravimetric calibrations
up to 20 liters or five gallons became
much faster and easier. NBS Handbook
145, SOP 14, Recommended Standard
Operations Procedure for Gravimetric
Calibration of Volume Ware Using an
Electronic Balance [7], was a practical
procedure for gravimetric volume calibration. Any laboratory with an adequate
electronic balance, appropriate mass
standards, a source of distilled or deionized water, and a metrologist trained in
precision mass measurement, could calibrate volume standards gravimetrically.
A complete set of calibrated metric
(15 milliliters to five liters) and avoirdupois (120 minims to one gallon) glass
standards were issued to each state laboratory by NBS in the 1960’s. These were
automatic burettes and pipettes primarily
www.ncsli.org
REVIEW PAPERS
-0.30
-0.35
Error (Cubic Inches)
-0.40
-0.45
-0.50
-0.55
-0.60
-0.65
-0.70
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Year
F i g u re 3. Gravimetric Calibration of North Carolina 5-Gallon Working Standard.
used to test glass standards for regulatory Weights and Measures field inspectors. Regional interlaboratory comparisons using both volume transfer and
gravimetric calibration comparisons
were completed on glassware and fivegallon test measures. Laboratory Auditing Problem (LAP) 28 was initiated to
monitor volume calibration capabilities
within the state laboratories. It was at
this time that the North Carolina Standards Laboratory started using gravimetric calibration for all volume calibrations
for standards up to four liters, or one
gallon. There were several reasons for
this change. There were significant safety
concerns with cleaning the glassware
standards with concentrated sulfuric
acid, which was drawn into the standard
with vacuum. In addition, the glass
volume standards used by the lab were
beginning to show signs of age with
chipped tips and drain times different
from the original NBS calibrations. Since
these aging volume standards had been
calibrated at NBS gravimetrically, gravimetric calibration eliminated a step in
the traceability chain for customer glassware standards. Changing to gravimetric
calibration avoided these problems.
Also, expense and risk of shipping
damage were decreased by not having to
send volume standards back to NIST for
calibration.
A 1990 regional interlaboratory comparison revealed a 0.4 cubic inch bias in
five-gallon calibrations made by the
North Carolina Standards Laboratory.
An internal gravimetric calibration elimVol. 1 No. 4 • December 2006
inated this bias. From that time on, the
one gallon, five-gallon and 20-liter
slicker plate reference standards in the
NC laboratory have been calibrated
gravimetrically. The accuracy and
repeatability of these calibrations have
been verified through years of history
(see Fig. 3) and multiple interlaboratory
comparisons.
8. G r a v i m e t r i c Wa t e r D r a w
Concept
In 1997, the NC 100-gallon stainless
steel prover was recalibrated by NIST.
One of the NC metrologists delivered the
prover and assisted John Houser with the
gravimetric calibration. From that point,
NC began planning for large volume
gravimetric calibration. However, it was
requests for SVP calibrations that were
actually the driving force behind the
development of our large volume gravimetric calibration procedure. With
increasing number of requests for SVP
calibrations, the NIST Weights and Measures Division decided to work with
North Carolina, Arizona, and Michigan
to establish gravimetric SVP calibration
capabilities on a regional level. The collaboration between the NIST Weights
and Measures Division and the Arizona,
Michigan, and North Carolina State
metrology laboratories, has answered
many questions and facilitated significant improvements in gravimetric calibration procedures. The Maine State
metrology laboratory metrologist developed a program in parallel with this initiative. Indiana Weights and Measures
Laboratory metrologists also calibrate
SVPs but use a procedure based on
volume transfer water draw into multiple
five-gallon test measures. Much has been
learned about the limiting factors in
gravimetric calibration as these programs have developed.
The water draw procedure, as
described by API in the Manual of Petroleum Measurement Standards [8], is a
volume transfer procedure. However, the
procedure lends itself well to gravimetric
calibration. The objectives of revising the
API procedure included:
• To minimize measurement uncertainties by limiting the procedure to the
actual measurement of the mass of the
water delivered by the vessel, rather
than subtracting the measured mass of
the "wet" vessel from that of the filled
vessel.
• To minimize the effect of balance nonlinearity by testing the balance sensitivity at the nominal mass of water
being weighed at the time of calibration.
• To minimize the effect of balance drift
by allowing the balance to be zeroed
immediately before each measurement.
• To minimize the effect of temperature
changes by shortening the duration of
the procedure.
P ro c e d u r e
The nine general steps of the weighing
procedure are summarized below.
9 .1 De t e r m in e St a n d ar ds t o Us e
Based on the nominal volume of the
prover to be calibrated (hereafter
referred to as unit under test or UUT),
calculate the approximate mass of that
volume of water. Choose mass standards
to use so that the mass of the standards
approximately equals the calculated
water mass. The difference between the
calculated water mass and the mass of
the standards should not exceed 0.05 %
of the calculated water mass. The mass
of this combination of standards will be
referred to as Ms.
For non-metric volumes, it may be
more efficient to add tare weights (tx) to
the calculated mass of the water so that
the mass of water plus tx equals a conMEASURE
|
65
REVIEW PAPERS
venient combination of mass standards
for Ms. Effort should be made to both
minimize the number of weights that
must be handled and make sure that Ms
is as close to the calculated water mass as
practical.
9 .5 R e c o r d A m b i e n t C o n d i t i o n s
f o r A i r D e ns it y C a l c ul a t io n
Ad vant ages
Record the air temperature, barometric
pressure, and relative humidity readings
to use for air density calculation.
In a traditional gravimetric calibration,
the unit under test is weighed drained
and then filled. To establish traceability,
each weighing is a comparison to standards with mass values nominally equal
to the weight of the wet down standard
and then the standard filled with water.
The new single pan electronic mass
comparators make this much easier than
the old equal arm balances, but it still
can be very awkward to move and position the prover being tested on the
balance pan. In the case of SVPs, it is
impossible. Filling the standard while it
is on the balance also introduces chances
for error and damage to the balance from
spilled water. In addition, the combinations of weights needed to nominally
equal the mass of the prover being tested
and then the filled prover can only be
determined after weighing the standard.
Often, the necessary combinations of
weights are cumbersome to use.
In a volume transfer water draw, water
is drained from the UUT into a standard
or series of standards equal to the
nominal volume. If this requires multiple
standards, the drain time of the prover
being tested must be interrupted each
time a standard is filled and replaced. In
addition to distorting the drain time, the
multiple transfers add opportunities for
mistakes. Since SVPs are displacement
provers, they have to be water drawn.
They also have volumes different from
typical standard provers. In order to calibrate a variety of SVPs with a single
drop, a laboratory would have to have
15, 20, 30, 40, 65, 75, 120, and 170gallon provers. It would be very difficult
for most labs to purchase and maintain
such an inventory of provers.
A gravimetric water draw can take
advantage of the best features of each
process. The standard provers can be
replaced with a simple transfer vessel
that is large enough to contain the full
volume of the prover being calibrated.
This vessel only serves to contain and
move water from the unit under test to
the balance and to the drain. Therefore,
drain characteristics or the physical
properties of the material are not important. Neither does it have to be the exact
volume of the prover being tested. One
9 .6 Pr e p a r e t h e U U T
9.2 We i g h t h e Tr a n s f e r Ve s s e l
to Det er mi n e Appro pr i a t e
Ta re We i g h t s
To establish the wet down condition of
the transfer vessel, fill it with tempered
distilled or deionized water above the
lowest point of the drain hole. Open the
drain and let this water drain out. Close
the drain and place the transfer vessel on
balance. This reading indicates the
approximate mass that will need to be
added to the balance and tared off
before the mass standards are placed on
the balance.
This mass will be referred to as tare
zero (t0). The t0 tare weight is used to
duplicate the approximate weight of the
transfer vessel. Like the transfer vessel,
the t0 weights will be zeroed off and
duplicate the effect of zeroing off the
transfer vessel. They will not be used in
any calculations. This step insures that
the mass of the water and mass of the
standards are both measured within the
same sensitivity range on the balance.
This avoids sensitivity uncertainties that
could be caused by balance nonlinearity
if the standard and the water were
weighed in different parts of the balance
electronic weighing range.
9 .3 F i l l t h e U U T i n P r e p a r a t i o n
f or C al i br at i o n
Fill the UUT to the nominal volume indication with distilled or deionized water
that is near ambient room temperature.
9.4 We i g h t h e M a s s S t a n d a rd s
fo r Readi ng O1
Place the tare zero (t0) weights on the
balance. Zero the balance. Place predetermined mass standards closely approximating the nominal mass of the water
volume to be measured on the balance.
Record this balance reading as O 1.
Remove the mass standards from the
balance.
66
|
MEASURE
Take the temperature of water in the
UUT. Verify that the UUT is still filled to
the nominal indication and adjust if necessary. This step should be repeated
immediately before draining into the
transfer vessel.
9.7 Ta re Off Tr a n s f e r Ve s s e l
Place the transfer vessel and lid on the
balance. Zero the balance.
9 .8 W e i g h W a t e r f o r R e a d i n g O 2
Immediately after zeroing the balance,
remove the transfer vessel and take it to
the unit under test. Drain the water from
the UUT into the transfer vessel. Drain
for 30 seconds after cessation of main
flow unless another drain time is specified. Place the lid on the transfer vessel
to limit evaporation. Carefully return the
transfer vessel to the balance. If needed,
add known tare weights (tx) to the
balance pan. This balance reading is
recorded as O2.
9 .9 R e p e a t P ro c e d u re
Repeat procedure until confidence has
been established in the measurement. All
repeated runs must agree within ±0.02 %.
Determine the mean of the results.
It should be noted that this revised
procedure primarily applies to vessels
calibrated “to deliver”. When a vessel is
calibrated “to deliver”, the interior is
wetted down with the liquid to be measured by a standard procedure immediately before each use. Typically, for
small, hand held measures, this “wet
down” requires a uniform 30-second
pour and a 10-second drain after it is
empty. For larger vessels, the filled vessel
is allowed to drain at full flow and then
allowed to drain an additional 30
seconds after cessation of main flow. The
advantage of calibrating a vessel to
deliver is that it can be used without
drying each time between uses. In order
to calibrate the vessel “to contain” using
this procedure, the dried vessel must be
used as the transfer vessel.
www.ncsli.org
REVIEW PAPERS
11. Wa t e r P u r i t y
F i g u re 4. Transfer vessels, pallet stacker, and mass comparator.
transfer vessel will work for a range of
calibrations. It is just important for the
transfer vessel to be sturdy, not leak,
have a lid to limit evaporation, and to be
stable when placed on the balance pan.
There also has to be an efficient way to
transport it to and from the balance. The
North Carolina Standards Laboratory
found two inexpensive, plastic, horizontal agricultural tanks that work well for
this job (see Fig. 4). One is 35 gallons for
provers between 15 and 30 gallons. The
other is 125 gallons for larger provers up
to 100 gallons. These tanks were banded
on wooden pallets so that they can be
moved with a pallet stacker. Other laboratories have a series of hooks and cables
to lift and move the tanks with an overhead crane. This seems to minimize
water movement within the tank and
improve balance stability. Regardless of
the method of transport, it is important
that the movement be smooth to mini-
mize sloshing of the product. In addition,
there must be adequate control to slowly
lower the vessel and center it on the
balance pan to avoid damage to the mass
comparator.
Another major factor in the accuracy
and repeatability of any gravimetric procedure is timing. The procedure has to be
completed before there are any significant changes in the water conditions
from when the unit under test is filled
and adjusted to nominal. It is also important that conditions affecting the balance
do not change between weighing the
mass standards and weighing the water.
The procedure should be carried out in a
well-controlled environment. NIST
Handbook 143 lists the requirements in
Table 1 for gravimetric volume calibration. [9]
More will be said about the ambient
environment in the water temperature
measurement section.
Procedure
Temperature
I
Gravimetric
20 °C to 23 °C, set point ± 2 °C
Maximum change 1.0 °C/hr.
Relative Humidity
(Maximum Range
per 4 hours)
40 % to 60 %
(50 % ± 10 %)
Ta bl e 1. Requirements for gravimetric volume calibration from NIST Handbook 143.
The purity of water in the traditional
volume transfer volume calibration is of
little consequence. Water needs to be
clean, but not necessarily pure. Of
course, these relative terms need definition. As a rule, potable water is clean
enough for volume transfer calibrations.
The purity of water is not as critical in
volume transfer since any uncertainty in
density measurement it causes, will be
applied equally to both the unit under
test and the standard itself. As long as
nothing is done to change the density of
the water between vessels (such as a
major temperature change), density
uncertainties cancel each other out.
Excessive air trapped in the water is a
problem in volume transfer calibration
however since it bubbles out over time.
Traditional gravimetric procedures
such as NIST Handbook 145, SOP 14,
specify ASTM D 1193 Type IV water. In
gravimetric calibration, the water density
is critical in two ways. First, an accurate
water density is necessary to calculate
the mass of the water. There is a significant difference between the density of
the mass standards (typically between
7.3 grams per cubic centimeter and 8.0
grams per cubic centimeter) and water
(approximately 1.0 gram per cubic centimeter). Therefore, the air buoyancy
correction is very significant to the mass
calculation. Second, the water density is
the only factor used to convert from
measured mass to volume. Regardless of
the accuracy and precision of the mass
determination, the volume conversion is
proportionally dependant on the accuracy of the water density. A 1 % uncertainty in water density will result in an
approximate 1 % uncertainty in volume.
A 0.001 % uncertainty in water density
will result in a one milliliter uncertainty
per 100 liters.
There are two ways to determine the
water density. The first is to measure it
directly. Commercial density meters are
available that read out to five or six
decimal places. Second, if an adequate
density meter is not available, water
density can be estimated based on an
assumption of water purity and precise
measurement of the temperature. Work
by Gary Cohrs at Calibron Systems, Inc.
[10] using Archimedes’ principle and a
MEASURE
|
67
REVIEW PAPERS
500 Gal lo n P ro v er Wa t e r Te m p e r a t u re D a t a
F irs t 60 M in u te s
Temperature (degrees C)
17.50
17.00
16.50
Mid Scale Nec
16.00
Top of Cylinde
Middle of Cylin
Bottom of Cyli
Ambient
15.50
0
10
20
30
40
50
60
Minutes
F i g u re 5. DI water filter, resistivity meter,
and pump.
500 Gal lo n P ro v er Wa t e r Te m p e r a t u re
F irs t 5 Da ys
17.50
68
|
MEASURE
21.50
vibrating tube density meter, led him to
conclude that either deionized (DI) or
17.00
distilled water should be used for gravimetric prover calibration. Based on the
assumption that the 16.50
only pollutants in
tap water meeting federal drinking standards that have a measurable affect on
density are dissolved16.00
salts, Mr. Cohrs
developed a relationship between city tap
water density and conductivity. However,
this estimation has not
yet been repro15.50
0
10
ducible in other laboratories and adds an
uncertainty factor that can be avoided by
using either DI or distilled water. The
North Carolina Standards Laboratory
purchased the DI water filtration system
shown in Fig. 5.
Using a resistivity meter immediately
after DI resin beds, water with a resistivity of 18,000,000 ohm (18-megohm) was
easily achievable. However, the resistivity rapidly degraded when put in a storage tank (a resistivity of 2-megohm
immediately after storage) or when piped
through schedule 80 PVC piping (6 to 7megohm). This 18-megohm water had
not been exposed to any significant
sources of contamination. Was the drop
in resistivity significant to the density of
the water?
To answer this question, Georgia
Harris of the NIST Weights and Measures Division collected water samples
from various labs around the country.
She tested all of these using a five-digit
23.50
19.50
Mid Scale Neck (800 cu in @ Zero)
Top of Cylinder
Middle of Cylinder
Mid Scale Neck (100 cu in
Bottom of Cylinder
17.50
Top of Cylinder
Ambient
Middle of Cylinder
20
30
15.50
40
50
Bottom of Cylinder
60
Ambient
Minutes
0
1000
2000
3000
4000
5000
6000
7000
Minutes
F i g u re 6. Water temperature equilibration.
water density meter. Samples of fresh DI
water, distilled water, reverse osmosis filtered water (RO), deionized reverse
osmosis water (RODI), and tap water
samples were collected. Some of the
samples had been stored for significant
lengths of time and some were recycled
water that had been used in previous
gravimetric calibrations. Though analysis
of her data is still ongoing, initial analysis indicates that DI, distilled, and RODI
water from the state metrology laboratories in Arizona, Maine, Michigan, and
North Carolina, the NIST volume group,
and the NIST Weights and Measures
Division training labs are all equivalent.
There were no significant differences in
density, regardless of the age of the
sample. On the other hand, the densities
of RO water and tap water from the
same locations differed significantly. The
preliminary conclusion is that if DI or
distilled water is initially pure (verified
by a conductivity or resistivity meter)
and is stored properly, density does not
change with normal storage. The appropriate water density formulas can still be
used to estimate the density based on
temperature. Ms. Harris also determined
that reverse osmosis does not always
filter out enough of the contaminants for
the water to have a predictable density. If
www.ncsli.org
REVIEW PAPERS
F i g u re 7. 1000-gallon water storage tank.
RO water is used, it should also be
deionized.
12. Wa t e r Te m p e r a t u r e
M e a s u re m e n t
Accurate measurement of the water temperature is critical to gravimetric calibration. An uncertainty of 1 ºC at 24 ºC will
result in a calibration uncertainty of
0.02 %. It is relatively easy to purchase
a thermometer with an accuracy of
0.1 ºC or better, but this does not solve
the problem completely. There are two
other major factors to consider. These
are the stability and uniformity of the
temperature in the standard (gradients).
Additional precautions need to be taken
to limit these potential sources of uncertainty.
The stability of the water can be
increased by storing for an extended
period in the temperature-controlled
climate where the calibration will take
place. Storage capacity should include
enough water to complete the calibration. Temperature equilibration requires
much time. Any difference between the
ambient temperature and the water temperature results in temperature changes
during the calibration as the water seeks
equilibrium with room temperature.
The uniformity of the water temperature is also critical to the measurement.
To analyze this, a 500-gallon stainless
steel prover was filled with tap water.
The results are shown in Fig. 6. In this
test, water temperatures were measured
tank. If the flow of water is stopped, the
at four different levels in the prover.
pressure increases to a point that water
These were the center of the neck scale
bypasses the system back up to the
(the line with squares), top of the cylinstorage tank rather than circulating in
der, middle of the cylinder, and bottom of
the pump. Another option would be to
the cylinder (lines with triangles and diamake sure the pump used is a magnetic
monds). The dashed line is the ambient
drive pump so that when flow stops, the
room temperature, which is off scale on
motor is allowed to turn relatively freely
the graph covering just the first hour. All
without heating up. Finally, if the storage
measurements were taken in the centertank is not at the same level or in the
line of the provers, straight down
same location as the prover calibration,
through the neck opening. Rapid
care must be taken so that the temperaresponse bare wire bead thermistors were
ture in the area of the tank is the same as
used as temperature sensors.
the calibration area. In North Carolina
Several things are clear from the data.
where the storage tank is on a warmer
It takes a considerable amount of time
mezzanine, fans were added to circulate
for the water temperature to equilibrate.
and mix the air properly. Figure 7 shows
The time would have been even longer if
the 1000-gallon storage tank with one of
there had been more of a difference
the circulating fans in the background.
between the water temperature and
ambient. In addition, as expected, the
neck temperature changes much faster
1 3. E q uat io n s
than the main body of water. This data
illustrates the importance of measuring
1 3 .1 W a t e r D e n s i t y
temperature in the center of the vessel
There have been several water density
and working quickly. In this example, the
tables and equations over the past few
fluid in the neck represented less than
years. Currently the formula used by
one percent of the prover volume. Since
NIST, and with the most international
the fluid level is gauged in the neck, it is
acceptance, is the Patterson/Morris 1994
important that it not be significantly difwater density, ρ(tw), equation as shown
ferent from the rest of the water. Again,
below. [11]
 
this points out the advantage of allowing
2
ρ (tw ) = ρ 1−  A(t − t (1)
) + B( t − t ) + C ( t
0
0
the water to reach temperature equilib0


rium with the room before calibrating
 
2
3
4
provers gravimetrically. As discussed
ρ (tw ) = ρ 1−  A(t − t ) + B( t − t ) + C ( t − t ) + D( t − t ) + E(
0
0
0
0
0


previously, in North Carolina, a 1000gallon vertical plastic agricultural tank
 
2
3
4
5 
(tw ) = ρ to
1−store
− t ) + B(
− t ) + C ( t − t ) + D( t − t ) + E( t − t )  
was ρ
purchased
DIt water
 A(t water.
0
0
0
0
0 
0
 store well. Resistivity

did not seem to
increased rapidly during storage and
piping. This made it difficult to prove the
water was still of adequate quality.
Therefore, water is stored as city water
before it goes through the DI beds. This
temperature equilibrated water is
pumped directly from storage, through
the DI filters, into the vessel being calibrated. This way the water is at the peak
condition for both temperature and
deionization.
Two more temperature obstacles must
be overcome. Care must be taken so that
the pump used does not heat the water.
This issue can be resolved by increasing
the supply pipe diameter to the pump
and piping a bypass line back to the tank.
The bypass should be controlled by a
pressure valve where it returns to the
where:
ρ0 = 999.97358 kg/m3
A = 7.0134 x 10–8 (ºC)–1
B = 7.926504 x 10–6 (ºC)–2
C = – 7.575677 x 10–8 (ºC)–3
D = 7.314894 x 10–10 (ºC)–4
E = – 3.596458 x 10–12 (ºC)–5
t0 = 3.9818 ºC
tw = temperature of the water in ºC.
Though the differences between equations are small and usually insignificant
to the measurement uncertainty, there
are differences. In our global economy,
international acceptance and agreement
are essential.
MEASURE
|
69
REVIEW PAPERS
Typical Uncertainty Components for Gravimetric Water Draw
Variable
Description
Components
Uncertainty
Estimate
Type of
Distribution
Ma
Molar mass of the air
within laboratory:
28.963 5 x 10-3 kg/mol
Std. Dev.
Process, s(p)
34 ppm
Normal
Pooled Standard Deviation
p
Ambient barometric
pressure in Pascal
Water
Temperature
6.3 ppm
Rectangular
Thermometer Uncertainty
is 0.0052 ºC –
increase factor times 10 for
gradients
T
Ambient temperature
in Kelvin
Water
Density
5.5 ppm
Rectangular
NIST, Jones/Harris
1992,10 ppm
Air
Temperature
0.43 ppm
Rectangular
Doubled water temp
Barometric
Pressure
0.40 ppm
Rectangular
Specifications from ability
to measure pressure
Relative
Humidity
0.31 ppm
Rectangular
Specifications from ability
to measure relative
humidity
Mass
Calibrations
0.29 ppm
Normal
R
h
Universal gas constant:
8.314 510 J mol-1 K-1
Relative humidity in %
1.000 62
+ (3.14 x 10-8) p
+ (5.6 x 10-7) t2
f
t
ambient temperature
in degrees Celsius
ρsv
1 Pascal x exp
(AT2 + BT + C + D/T)
A
1.237 884 7 x 10-5 K-2
B
–1.912 131 6 x 10-2 K-1
C
33.937 110 47
103
–6.343 164 5 x
a0
1.581 23 x 10-6 K Pa-1
a1
a2
–2.933 1 x
34.9 ppm Combined Uncertainty (k = 1)
69.8 ppm Expanded Uncertainty (k = 2)
K
Pa-1
1.104 3
x 10-10 K-1 Pa-1
b0
5.707 x 10-6 K Pa-1
b1
–2.051 x 10-8 Pa-1
c0
1.989 8 x 10-4 K Pa-1
c1
–2.376 x 10-6 Pa-1
1 3 . 2 A i r D e ns i ty
The air density, ρ, should be calculated
using the International Committee for
Weights and Measures (CIPM) air density
equation formula. [12]
pMa
ρ =
ZRT
(1 −
0.3780 xv )
|
MEASURE
Compute the volume, Vt , for each determination using the equation:
   M  
 
 ρ    1
s
a
1 – ρa   – M(3)
Vt =   O2 
t 1–

 

ρs  
 ρt    ρw – ρa
   O1  
(
(d +
)
exv2
Variables for CIPM air density equation
are shown in Table 2.
Ta bl e 2. Variables for CIPM air density
equation.
13.3 Vo l u m e a t C a l i b r a t i o n Wa t e r
Te m p e r a t u re ( T w )
ρ
 
 ρ    1 
 Ms  
xv = ( h / 100 ) f psv=   O (2)
1 – ρa   – M t 1– a   
Vt   2 



ρs  
 O1  
 ρt    ρw – ρa 
 
p
p2
a + a t + a t 2 + ( b + b t ) x + (where:
Z =1−
c0 + c1t ) xv2  + 2 d + exv2
0
1
2
0
1
v
T
O1 = ObservationT#1, balance reading
p
p2
-11 2 2 Pa-2
a 1.83
Z = 1d −
+ a1tx 10
+ a2t K
+ ( b0 + b1t ) xv + ( c0 + c1t ) xv2  + 2
0
T
T
e
–0.765 x 10-8 K2 Pa-2
70
Mass uncertainty from
standards calibration
report (k = 1)
Ta bl e 3. Uncertainty components of gravimetric water draw to be included in the
uncertainty statement.
D
10-8
Comments
)
for mass standard
O2 = Observation #2, balance reading
for water delivered from vessel
Ms = Mass of mass standard(s)
Mt = Mass of tare weight(s)
ρa = Air density
ρs = Density of mass standard(s)
ρt = Density of tare weight(s)
ρw = Water density
www.ncsli.org
REVIEW PAPERS
Additional Uncertainty Components for Small Volume Provers
Uncertainty
Type of
Estimate
Distribution
Water
Compressibility
1.8 ppm
Rectangular
NIST, Jones/Harris 1992, 10 ppm
Prover Temperature
2.5 ppm
Rectangular
Temperature from thermometer well. Transference between water
and metal is not good – double the thermometer uncertainty
Detector
Temperature
0.17 ppm
Rectangular
Temperature from bar. Transference between water and metal is
not good - double the thermometer uncertainty
Prover Pressure
0.025 ppm
Rectangular
Uncertainty @ 60 psi from calibration certificate
Area Thermal
Expansion (Ga)
0.58 ppm
Rectangular
Estimate @ 1 % of number given by calibration report, cubical
coefficient is 2 %
Detector Thermal
Expansion (Gl)
0.051 ppm
Rectangular
coefficient is 2 %
Mod. of Elasticity (E)
0.089 ppm
Rectangular
coefficient is 2 %
0.00074 ppm
Rectangular
Estimate accuracy to last number in value by one division
0.0010 ppm
Rectangular
Estimate accuracy to last number in value by one division
Components
Flow Tube Diameter
(ID)
Tube Wall Thickness
(WT)
Comments
35.0 ppm Combined Uncertainty (k = 1)
70.0 ppm Expanded Uncertainty (k = 2)
Ta bl e 4. Additional uncertainty components for small volume provers to be included in the uncertainty statement.
13.4 Vo l u m e a t a R e f e r e n c e
Te m p e r a t u re
1 3. 4. 2 Co m put ati o n o f V 60
Compute the volume at 60 °F, V60 , for
each determination using the expression:
1 3. 4. 1 Co m put at io n o f V20
Compute the volume at 20 °C, V20 , for
each determination using the expression:
V20 = Vt [1 – α (t w – 20)]
(4)
where α is the cubical coefficient of
expansion (1/°C) of the container being
calibrated, and tw is the temperature of
the water for each determination (in °C).
Repeat for each determination and calculate the mean, l
V20, for the duplicate
measurements.
V60 = Vt [1 – α (t w – 60)]
(5)
where α is the cubical coefficient of
expansion (1/°F) of the container being
calibrated, and tw is the temperature of
the water for each determination (in °F).
Repeat for each determination and calculate the mean, l
V60, for the duplicate
measurements.
14. Unc er tai nty Fa ctor s
Traceability is not complete without an
uncertainty statement. Working with
Georgia Harris (NIST), Kelleen Larson
(Arizona), Craig VanBuren (Michigan),
and Bill Erickson (Michigan), the components of gravimetric water draw
uncertainty factors listed in Table 3 have
been identified.
Additional factors that need to be considered for a small volume prover calibration are listed in Table 4. Most of
these are based on information specified
by the SVP manufacturer and are brand
and model specific.
One additional uncertainty factor has
to be considered for the readability of the
scale plate for open neck provers. This is
estimated to be approximately one tenth
of the smallest scale graduation. These
are just typical numbers. In our experience of calibrating 27 small volume
MEASURE
|
71
REVIEW PAPERS
provers (five runs each) and 11 open
neck provers (ten runs each), our best
standard deviation of the process was 10
ppm and the worst was 93 ppm with a
median value of 22 for traditional open
neck provers and 38 ppm for small
volume provers. This variance is to be
expected due to technique, drain characteristics, seal and valve integrity, and a
multitude of other factors that make
each standard a unique device.
15. C o n c l u s i o n s a n d L o o k i n g
t o t h e F u t u re
Based on this work, gravimetric water
draw is a viable volume calibration
option. It has the advantages of simplicity, realistic drain times, speed, and relatively low additional equipment outlay
for the mass calibration laboratory.
This is a new area for the State Weights
and Measures laboratories. Work will
continue in order to build additional
long-term degrees of freedom into the
process uncertainty analysis. A 100gallon control standard with a significant
amount of volume transfer calibration
history is being calibrated by gravimetric
water draw now to compare the two
methods directly. The calibration process
will continue to be refined. Additional
data will be collected and input from
other laboratories will be solicited in
order to continue to improve large
volume calibration and subsequently,
enhance meter proving capabilities in the
United States.
16. A c kno wl e dg em e nts
I thank the following: (1) The metrologists of the North Carolina Standards
Laboratory, Bob Albright, Tal Anderson,
Van Hyder, Cliff Murray, and Sharon
Woodard who, in addition to working on
the calibration procedure itself, made the
hundreds of measurements referenced in
this paper. (2) Henry Oppermann,
Georgia Harris, and Val Miller of the
NIST Weights and Measures Division,
for providing support for travel and
training, measurement evaluation, document review, and general wisdom. (3)
Calibron Systems, Inc., Scottsdale,
Arizona, for their technical expertise,
SVP calibration spreadsheet that served
as a starting point for ours, and for inviting several of our group visit their facil72
|
MEASURE
ity to observe their calibration of a SVP.
(4) Kelleen Larson from the Arizona
Department of Weights and Measures,
and Craig VanBuren and Bill Erickson
from the Michigan Department of Agriculture, E.C. Heffron Metrology Laboratory, for the hours of work developing,
revising, and perfecting procedures. (5)
The loss control specialists and SVP
operators of Marathon Ashland Petroleum, LLC, for their incredible patience
as we learned by experimenting with
their SVPs.
17 . R e f e re n c e s
[1] Energy Information Administration,
Office of Oil and Gas, U.S. Department
of Energy, Washington, DC 20585,
Petroleum Marketing Monthly, March
2005. [Available from the web site
eia.doe.gov/pub/oil_gas/petroleum/data
_publications/petroleum_marketing_mo
nthly/historical/2005/2005_03/pmm_2
005_03.html]
[2] NIST Handbook 44, “Specifications,
Tolerances, and Other Technical
Requirements for Weighing and Measuring Devices,” Table T.2, pp. 3-15, 2005
Edition. [Available from the website
http://ts.nist.gov/ts/htdocs/230/235/pu
bs.htm]
[3] NIST Handbook 105-3, “Specifications
and Tolerances for Reference Standards
and Field Standard Weights and Measures, Graduated Neck Type Volumetric
Field Standards,” 1997 revision. [Available from the website http://ts.nist.gov/
ts/htdocs/230/235/pubs.htm]
[4] American Petroleum Institute, “Manual
of Petroleum Measurement Standards,”
Chapter 4: Proving Systems, Section 7:
Field Standard Test Measures, 2nd
Edition, December 1998.
Small Volume Provers, July 1988.
Pulse Interpolation, November 2003.
[7] J.K. Taylor and H.V. Oppermann,
“Handbook for the Quality Assurance of
Metrological Measurements,” NBS
Handbook 145, pp. SOP 14-1 to SOP
14-4, 1986. [Note: Selected updates are
contained in G. L. Harris, and J. A.
[8]
[9]
[10]
[11]
[12]
Torres, “Selected Laboratory and Measurement Practices and Procedures to
Support Basic Mass Calibrations,” NIST
IR 6969, March 01, 2003.]
American Petroleum Institute, “Manual
Chapter 12: Calculation of Petroleum
Quantities, Section 2: Calculation of
Petroleum Quantities Using Dynamic
Measurement Methods and Volumetric
Correction Factors, Part 4: Calculation
of Base Prover Volumes by Waterdraw
Method, 1st Edition, December 1997,
Reaffirmed March 2002.
NIST Handbook 143, “State NIST
Weights and Measures Laboratories
Program Handbook,” Table 10, p. 66,
March 2003. [Available from the
website http://ts.nist.gov/ts/htdocs/230/
235/pubs.htm]
Gary Cohrs, “Water Density vs. Conductivity Using Archimedes’ Law for
Density Measurement,” Handbook for
the Quality Assurance of Metrological
Measurements, Presentation for API
4.9.4 Committee Meeting, February
2004.
J.B. Patterson and E.C. Morris, “Measurement of Absolute Water Density, 1
°C to 40 °C,” Metrologia, vol. 31, pp.
277-288, 1994.
P. Giacomo, “Equation for the Determination of the Density of Moist Air,”
Metrologia, vol. 18, pp. 33-40 (1982)
and R.S. Davis, “Equation for the Determination of the Density of Moist Air,”
Metrologia, vol. 29, pp. 67-70 (1992).
www.ncsli.org
74
|
MEASURE
www.ncsli.org
NEW PRODUCTS
NEW PRODUCTS
Sypris Test & Measurement,
Inc. Adds New Calibrations
Sypris Test & Measurement, Inc., a subsidiary of Sypris Solutions, Inc., has
added Antenna, Isotropic/Electromagnetic Field Probe, Line Impedance Stabilization Network calibrations and RF
Screen Room/Anechoic Chamber Certification to its suite of calibration service
offerings. Current calibration services
include AC/DC, RF, Physical Dimensional, and Temperature instruments and
test equipment.
For more information, 800-463-8786 or
www.calibrationandrepair.com.
Veriteq Announces Calibration
Lab Monitoring System
Veriteq Instruments, Inc. announces the
release of viewLinc 3.0 enabling calibration labs to monitor critical applications
and processes. Personnel can view and
monitor temperature and RH environmental data locally or across an existing
network with a standard web-browser.
With multi-stage alarm notifications to
cell phones, pagers and PCs, and simple
installation, viewLinc 3.0 with Veriteq
data recorders allows a faster response to
fluctuating conditions before critical calibration
processes
are
affected.
viewLinc’s scalability enables viewing
historical or real-time data from one to
100 loggers. Veriteq recorders store data
internally, archive data automatically
and provide temperature and RH accuracy to `/– 2 %RH and `/–0.15 °C,
and A2LA accredited calibration.
For more information, see www.veriteq.com.
METRICA Opens Subsidiary
in San Antonio, TX
Metrica, S.A. de C.V., a company
founded in 1992 and headquartered in
Monterrey, Mexico, has recently opened
a subsidiary in San Antonio, TX named
Metrica Industries Co., which is designed
to provide calibration services for industrial measuring instruments in the areas
of mass, pressure and electrical. Metrica
Industries Co. offers a wide variety of
metrology services to support the ISO
9000 and TS 16949 Quality Standards,
including: calibration of measuring
instruments; fabrication of special measVol. 1 No. 4 • December 2006
uring devices used in the automotive
industry; technical consulting; sale of
industrial measuring instruments; equipment maintenance; training; analysis for
measuring instruments; and, the most
important of all, offering customized
solutions for their clients. The main calibration laboratory is located in Monterrey,
Nuevo Leon, Mexico with several mobile
units designed to provide “on site” calibration services for each of their clients.
For more information, please contact Mr.
Moisés Rivera at [email protected] or
Mr. Roberto Benitez at [email protected].
ACLASS Signs APLAC MRA
At the 12th Annual General Assembly in
Taipei, Chinese Taipei, ACLASS signed
the Asia Pacific Laboratory Accreditation Cooperation (APLAC) Mutual
Recognition Arrangement (MRA) on
September 13, 2006 for both calibration
and testing. APLAC groups accreditation
bodies in the Asia Pacific region that are
responsible for accrediting calibration,
testing and inspection facilities. As a
result of signing the APLAC MRA,
ACLASS, already an associate member
in ILAC, now automatically progresses
to full membership of ILAC, in the fields
of testing and calibration. A formal
signing ceremony for the ILAC MRA will
take place November 12, 2006 in
Cancun, Mexico. ACLASS is headquartered in the Washington, DC metro area
and is an accreditation body for the
ISO/IEC 17025 standard.
For more information, www.aclasscorp.com.
A2LA Adds Inspection Body
Accreditation to its APLAC
Scope of Recognition
A2LA has successfully completed the
Asia Pacific Laboratory Accreditation
Cooperation (APLAC) peer re-evaluation process. A2LA recognition for
testing and calibration has been extended
for four years. In addition, this year,
inspection body accreditation to
ISO/IEC 17020 and ILAC/IAF A4 has
been added to its scope of recognition.
“A2LA’s addition of inspection body
accreditation to our MRA signatory
status gives the program credibility and
instant worldwide recognition,” said
Peter Unger, President of A2LA. “A2LA
accredited inspection bodies can now
benefit from the same international
recognition and credibility that our
testing and calibration laboratories have
enjoyed for years.”
For more information, www.A2LA.org.
DeFelsko Redesigns
PosiTector 200-Advanced
Coating Thickness Gage
The PosiTector 200
coating thickness gage
has
just
been
redesigned to include
new features, new
measurement ranges
and new models. The updated PosiTector
200 is ideal for measuring the total thickness of a coating system or up to 3 individual layer thicknesses in a multi-layer
system. New features include extended
range probes, faster measuring speed,
large impact-resistant Lexan® display,
IP5X ingress protection and protective
rubber holster. Conforms to ASTM
D6132 and ISO 2808.
For more information, (800) 448-3835 or
www.defelsko.com.
Fluke Offers 8845A/8846A
Precision Digital Multimeters
The new Fluke
8845A and 8846A
Precision Digital
Multimeters
feature 6.5 digit
resolution, have a dual display that
shows data in graphic or numeric formats,
and provide multifunction measurement
capability. They have 14 measurement
functions, extending the capability of a
standard DMM with wider ranges and
features to measure temperature, capacitance, period and frequency. The 2 x 4
ohms function uses patented split terminal jacks that allow users to perform 4wire measurements using only two leads
instead of four. The meters measure
volts dc with an accuracy of up to 0.0024
%, have a voltage range of 100 mV to
1000 V with up to 100 nV resolution,
current range of 100 µA to 10 A with up
to 100 pA resolution, and have a wide
ohms measurement range from 10 Q to 1
GQ with up to 10 µQ resolution.
For more information, (888) 308-5277, or
www.fluke.com/884XA.
MEASURE
|
75
76
|
MEASURE
www.ncsli.org
NEW PRODUCTS
Guildline Introduces New
Series Of DCC bridges
Hart Scientific Announces
Temperature/Humidity Logger
Guildline Instruments Ltd. recently
launched the 6622A Series of Direct
Current Comparator Bridges for resistance and thermometry applications.
There are five models in the 6622A
series, each with different measurement range and accuracy. Best uncertainty is 0.04 ppm and the widest range
from 1mQ to 1GQ. These bridges are
fully upgradeable in measurement
uncertainties and ranges. A measurement ratio of 100:1 is available on all
models, contributing to lower uncertainties when laddering up/down, reducing
the number of standards a lab needs and
making it possible to verify the bridge
performance without a Hamon style
transfer standard.
Optional range
extenders can expand the measurement
range down to 0.1 µQ at up to 4200A
maximum current.
Hart Scientific has
announced the new
Model
1620A,
“DewK,” paperless
temperature
and
humidity data logger with wireless, Ethernet, and RS-232
communications. Designed to facilitate
the electronic management of environmental temperature and humidity data,
this thermo-hygrometer is intended for
critical locations such as calibration and
research labs, pharmaceutical and
chemical storage areas, and many
medical environments. The DewK
accepts inputs from up to two sensors,
which may be mounted directly on the
unit or up to 30 meters away. Two
sensor models are available.
For more information, www.guildline.com.
For more information, www.hartscientific.com.
Integrated Sciences Group
Metrology Handbook
The Analytical Metrology Handbook,
introduced at the 2006 NCSLI Workshop & Symposium, is available for $25
from Integrated Sciences Group at
isgmax.com. Topics include measurement uncertainty analysis, calibration interval
analysis, measurement
decision risk analysis
and SPC for measurement processes.
Order on-line, or call
1-800-400-7866.
Keithley Metrology Achieves
ISO 17025 Accreditation
Keithley Instruments, Inc. announced
that its Metrology Services Group has
been
accredited
to
ISO/IEC
17025:2005 by the American Association for Laboratory Accreditation
(A2LA). This accreditation recognizes
that Keithley’s Metrology Services meet
the requirements of this international
standard, demonstrating its technical
competence to carry out very high-level
calibrations that are essential for many
of Keithley’s instruments, which have
MEASURE
|
77
NEW PRODUCTS
measurement resolutions below 1 femtoamp. The scope of accreditation can be
viewed at
www.a2la.org/scopepdf/
2462-01.pdf. Keithley’s customers can now
obtain accredited calibrations for Series
2300 sources, Series 2600 and Series 2400
SourceMeter® instruments, Integra Series
products, Series 2000 DMMs, and the
Model 6517A Electrometer.
For more information,www.keithley.com/pr/057.
Mettler Toledo Introduces
Two New Comparators
In metrology, comparators are used to determine the mass of weights
or samples to the highest
degree of accuracy. The
new XP56C Comparator
from Mettler Toledo has a
continuous
weighing
range of 52 g and a readability of 1 µg
(1/1000000 g), In combination with its
dedicated hanging weighing pan that
eliminates corner load errors, the XP56C
Comparator guarantees the world’s most
accurate and repeatable results. The
78
|
MEASURE
WeighCom application guides the user
step by step through the mass determination process, and SmartScreen, the color
touchscreen display, enables convenient
and error-free operation.
by Precision Measurements. Using state
of the art automated calibration equipment, Precision Measurements can also
repair and calibrate all thermal converters, independent of their manufacturer.
For more information,
[email protected].
For more information, www.measure-tech.com.
Precision Measurements
AC/DC Transfer Standard
A new process
for manufacturing
thermal converters
has been developed
by Precision Measurements. Each unit is now being manufactured with Evanohms’ heater wire
and cold bead, which results in lower
AC/DC and reversal errors. This unit has
been tested by national labs and proven
to have the lowest errors. The versatility
of the new process allows the use of platinum leads and Evanohm leads for the
heater wire, which results in extremely
flat response to 100 MHz and beyond.
Both the new vacuum thermocouple and
the thermal converter are manufactured
Stranaska Develops UV/VIS
Calibration Artifacts
Holmium oxide solutions, comprised of
4 % (w/v) holmium oxide in aqueous
10 % (v/v) perchloric acid, provide
traceability for the calibration of absorption spectrophotometer wavelength
scales within the spectral range of 240
nm to 650 nm. To mitigate extrapolated
wavelength calibrations outside this traditional wavelength range, Stranaska
has developed two different series of
wavelength standard artifacts which
facilitate science-based measurement
traceability for reference wavelength
assignments above 650 nm and below
240 nm. One series is comprised of new
calibration artifacts certified for wavelength assignments in the spectral range
200 nm to 240 nm, and a second series is
www.ncsli.org
NEW PRODUCTS
certified for wavelength assignments in
the spectral range 650 nm to 900 nm.
Tegam Introduces RF Power
Calibration System IIB
Tovey Engineering Offers
Torque Calibration Systems
For more information, 970-282-3840,
www.stranaska.com.
The new Tegam
System IIB is an
updated version
of the Tegam/
Weinschel System IIA. The new System
IIB offers better functionality, fewer
components and a lower price. The
central component of the System IIB is
the new 1806A Dual Type IV Power
Meter, which allows for more precise RF
power control when calibrating higher
frequency, multi-range, or high and low
power sensors. Tegam offers a SYSIIB
Selection Guide online to assist with
package selection based on power sensor
calibration needs.
Tovey Engineering
transfer standard
torque calibration
systems are now
available in capacities ranging from 200 in-lb to 100,000
in-lb. These fully automated systems
permit efficient in-house transducer calibration. Systems include: torque frame,
transfer standard load cells, hydraulic
actuator, and software for automated
control and data analysis. The systems
feature full automation, precision
machining, very high accuracy reference
standard load cells, reduced misalignment errors, efficient transition from
clockwise to counterclockwise calibration, and true zero deadband measurement.
Team Torque Inc. Announces
A2LA Accreditation
Team Torque Inc. announced the latest
accreditation of their Calibration Laboratory and Customer Support Center.
The American Association for Laboratory Accreditation (A2LA) awarded the
company with accreditations to ISO
17025 and Z540-1, the top international
standards for calibration labs. Team
Torque Inc. provides calibration and
repair for all torque tools and calibration
test equipment. This recognition is
designed to prove that these services are
conducted at the highest level of technical competency available in the industry.
For more information, contact Jim Mueller
at (701) 223-4552, ext. 21.
For more information, call Kim Niznik Goff,
at (440) 466-6100, visit www.tegam.com.
Product information is provided as a
reader service and does not constitute
endorsement by NCSLI. Contact
information is provided for each product
so that readers may make direct inquiries.
For more information, (623) 434-5110 or
see www.toveyengineering.com.
ADVERTISER INDEX
A g i l e n t Te c h n o l o g i e s www.agilent.com....................................................
18
T h e A m e r ic a n A s s o c ia t i o n f o r L a b o r a t o r y A c c re d it a t io n
I n t e g ra t e d S c i e n c e s G ro u p www.isgmax.com ........................................
20
L a b o r a t o r y A c c re d i t a t i o n B u re a u www.l-a-b.com..................................
74
( A 2L A ) www.a2la.org ............................................................................
9
L e a r n i n g M e a s u re www.learningmeasure.com ........................................
80
A nd e e n - H a g e rl i n g , I n c. www.andeen-hagerling.com ..............................
76
Ma s y S y s te ms , I nc . www.masy.com ........................................................
11
A s s e t S m a r t www.assetsmart.com ............................................................
31
M e n s o r C o r p o r a t i o n www.mensor.com ....................................................
4
B l u e M o u n t a i n Q u a l i ty R es o u rce s www.coolblue.com ..........................
16
M o reh o u s e I n s t r u m e nt C o . www.mhforce.com ......................................
30
C al La b www.callabmag.com ....................................................................
80
N o r t h ro p G r u m m a n C o r p o r a t i o n www.northropgrumman.com ............
13
D a t a P roo f www.dataproof.com ................................................................
76
Oh m - L a b s www.ohm-labs.com ................................................................
76
D H I n s t r u m e nt s , I n c . www.dhinstruments.com ......................................
6
D H - B u d en b e rg I n c . www.dh-budenberginc.com ....................................
21
Q u a m e t e c C o r p o r a t i o n www.quametec.com ........................................
21
D y n a m i c Te c h n o l o g y, I nc . www.dynamictechnology.com ......................
10
S p ek t r a www.spektra-usa.com..................................................................
80
E d i s o n M e t ro l o g y www.edisonmetrology.com ........................................
73
Sy pr i s Te s t an d M e as u re m e n t www.calibration.com ..............................
14
E s s c o C a l i b r a t i o n www.esscolab.com ....................................................
77
TA C / To u r A n d o v e r C o n t ro l s www.tac.com/pe ........................................
78
F l u ke / H a r t S c ie n t if ic C o r p o r a ti o n
www.fluke.com ........................................................
T h u n d e r S c i e n t if i c Co r p o r a ti o n
Inside Front Cover, 11
G u lf Ca li b ra t io n S y st e m s
www.gcscalibration.com ..........................................
Outside Back Cover
H e u s s e r N e w e i g h www.neweigh.com ......................................................
74
H o l t I n s t r u m e n t www.holtinstrument.com ................................................
76
P ro c e s s I n s tr u m e n t s I n c. www.procinst.com .......................................... 45
www.thunderscientific.com ............................................Inside Back Cover
To v e y E n g i n e e r i n g www.toveyengineering.com ......................................
17
Vai s a l a I nc . www.vaisala.com....................................................................
45
Ve r i t e q www.veriteq.com ..........................................................................
15
MEASURE
|
79
CLASSIFIEDS
NCSLI Training Center
The NCSLI Training Center is designed to
provide a state-of-the-art training facility
with full electronic and staff support.
NCSLI member organizations are charged
a reduced rate. The facility offers:
• 900 Square Feet of Meeting Space
• State-of-the art Audiovisual Equipment
• Access to Full Catering Services and Kitchen
• Holds 45 Theater; 30 Classroom; 24 Square;
and 20 U-Shaped
• Pricing (NCSLI Members):
1-2 day – $350/day; 3 day – $275/day
• Audiovisual: Screen;
Flip charts/markers;
Electronic board;
Overhead and LCD
projection.
• High Speed Internet
• Ample Free Parking
Located at: NCSL International Headquarters
2995 Wilderness Place, Suite 107 • Boulder, CO 80301
303-440-3339
80
|
MEASURE
Training and Consulting Services
Providing training & consulting services for
those in measurement and metrology
related fields. Course offerings designed
for engineers and technicians.
We are currently looking for course beta
testers and consultants to be part of our
consultant network.
Website:www.learningmeasure.com
measure
N C SL I N T E RN AT I O N A L
Advertising
Opportunities
NCS L INTE
RNAT
measure
measure
IONAL
NC SL
INT ERN
The Journa
l of Measu
remen
The Jou
rnal
AT I O N
AL
t Scien ce
of Mea
sure
VOL. 1 NO.
1 • FALL
2005
men t Scie
In This
nce
Issue:
Lorem ipsum
dolor sit amet,
consectetue
r adipiscing
elit.
Donec interd
um.
VOL.
1 NO.
1 • FALL
Nam loremIn This Issue:
augue, sollic
itudin
nec, Mauris
Lorcursu
s, ligula feugia
em ipsu
pharetra ultrice
t
m dolor
consecs in arcu.
sit ame
tetu
Donec inte er adipiscing t,
Pellentesqu
elit.
rdum.
e habitant
morbi
tristique senec
tus et netus
Nam
lore
et malesuada
m augue,
ac turpis
nec, Mau
soll
as.icitudin
ris curegest
sus
pharetr
a ultrices , ligula feugiat
in arcu.
Pellente
sque hab
tristiqu
itant mor
e sen
et malesu ectus et netu bi
s
ada ac
turpis ege
stas.
All advertising is reserved
exclusively for NCSLI Member
Organizations. For pricing, specifications,
publication calendar, and deadlines, see
www.ncsli.org/measure/ads.cfm or contact
Craig Gulka, 303-440-3339,
[email protected]
www.ncsli.org
2005
December 2006
measure
NCSL INTERNATIONAL
The Journal of Measurement Science
NCSL International
In This Issue:
measure • The Journal of Measurement Science
Gravimetric Calibration of
Volumetric Standards with
Requirements for the Calibration
of Measuring and Test Equipment
Vol. 1 No. 4

ANSI/NCSL Z540.3:2006

Transcription

Similar documents

Product brochure MSA PrimaX IR Gas Transmitter

GAGEtrak 7 Brochure

HERE

Tribotester - ASEC Kunststoffen

ColorFacts 7.5 C3 Information_EN

The Analysis Of Dioxins And Furans Using HRGC-High

Basic Procedures in Force Calibration at NIST

VP-EYE 6.600a - inShop.hu webáruház

Ensuring consistent quality of cult drink with appropriate calibration

TERX0195_Impulse Control Sys-2 w.indd