American National Standard Guide For Electromagnetic CompatibilityComputations and Treatment of Measurement Uncertainty

American National Standard
Guide for Electromagnetic

Compatibility—Computations and
Treatment of Measurement
Uncertainty
C63®
Sponsored by the
Accredited Standards Committee C63® — Electromagnetic Compatibility
Sponsored by the
Accredited by the American National Standards Institute
IEEE
3 Park Avenue ANSI C63.23-2012
New York, NY 10016-5997
USA
13 March 2013
Authorized licensed use limited to: UNIVERSIDAD AUTONOMA DE NUEVO LEON. Downloaded on June 06,2018 at 18:22:56 UTC from IEEE Xplore. Restrictions apply.
Uncertainty
Accredited Standards Committee C63® – Electromagnetic Compatibility

accredited by the
American National Standards Institute
Secretariat
Institute of Electrical and Electronics Engineers, Inc.
Approved 3 December 2012
American National Standards Institute
Abstract: Methods for estimating measurement uncertainty of emissions measurement results
are provided, for use in conjunction with the basic methods of ANSI C63.4. Included in this
document are both Type A and Type B uncertainty evaluation methods.
Keywords: ANSI C63.23, electromagnetic compatibility (EMC), emissions, immunity,
measurement uncertainty, metrology, Type A evaluation, Type B evaluation
The Institute of Electrical and Electronics Engineers, Inc.

3 Park Avenue, New York, NY 10016-5997, USA
Copyright © 2013 by The Institute of Electrical and Electronics Engineers, Inc.

All rights reserved. Published 13 March 2013. Printed in the United States of America.
IEEE is a registered trademark in the U.S. Patent & Trademark Office, owned by The Institute of Electrical and Electronics
Engineers, Incorporated.
PDF: ISBN 978-0-7381-8222-3 STD98135

Print: ISBN 978-0-7381-8251-3 STDPD98135
IEEE prohibits discrimination, harassment and bullying. For more information, visit http://www.ieee.org/web/aboutus/whatis/policies/p9-26.html.
No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written permission
of the publisher.
An American National Standard implies a consensus of those substantially concerned with its scope and
provisions. An American National Standard is intended as a guide to aid the manufacturer, the consumer,
and the general public. The existence of an American National Standard does not in any respect preclude
anyone, whether he has approved the standard or not, from manufacturing, marketing, purchasing, or using
products, processes, or procedures not conforming to the standard. American National Standards are
subject to periodic review and users are cautioned to obtain the latest editions.
CAUTION NOTICE: This American National Standard may be revised or withdrawn at any time. The
procedures of the American National Standards Institute require that action be taken to reaffirm, revise, or
withdraw this standard no later than five years from the date of publication. Purchasers of American
National Standards may receive current information on all standards by calling or writing the American
National Standards Institute.
Authorization to photocopy portions of any individual standard for internal or personal use is granted by
the Institute of Electrical and Electronics Engineers, Inc., provided that the appropriate fee is paid to
Copyright Clearance Center. To arrange for payment of licensing fee, please contact Copyright Clearance
Center, Customer Service, 222 Rosewood Drive, Danvers, MA 01923 USA; (978) 750-8400. Permission to
photocopy portions of any individual standard for educational classroom use can also be obtained through
the Copyright Clearance Center.
Errata
Users are encouraged to check the IEEE Errata URL (http://standards.ieee.org/findstds/errata/index.html),
and the one for ASC C63® at http://www.c63.org/explanations_interpretations_request.htm, for errata
periodically.
Interpretations
Current interpretations can be accessed at the following URLs: http://www.c63.org/explanations_
interpretations_request.htm.
For more information about the committee that produced and maintains this standard, visit the ANSI
Accredited Standards Committee C63® web site at http://www.c63.org.
Notice to users
Laws and regulations

Users of these documents should consult all applicable laws and regulations. Compliance with the
provisions of this guide does not imply compliance to any applicable regulatory requirements.
Implementers of the guide are responsible for observing or referring to the applicable regulatory
requirements. IEEE does not, by the publication of its standards, intend to urge action that is not in
compliance with applicable laws, and these documents may not be construed as doing so.
Copyrights
This document is copyrighted by the IEEE. It is made available for a wide variety of both public and
private uses. These include both use, by reference, in laws and regulations, and use in private self-
regulation, standardization, and the promotion of engineering practices and methods. By making this
document available for use and adoption by public authorities and private users, the IEEE does not waive
any rights in copyright to this document.
Updating of IEEE documents

Users of IEEE Standards documents should be aware that these documents may be superseded at any time
by the issuance of new editions or may be amended from time to time through the issuance of amendments,
corrigenda, or errata. An official IEEE document at any point in time consists of the current edition of the
document together with any amendments, corrigenda, or errata then in effect. In order to determine whether
a given document is the current edition and whether it has been amended through the issuance of
amendments, corrigenda, or errata, visit the IEEE-SA Website at http://standards.ieee.org/index.html or
contact the IEEE at the address listed previously. For more information about the IEEE Standards
Association or the IEEE standards development process, visit the IEEE-SA Website at
http://standards.ieee.org/index.html.
Errata
Errata, if any, for this and all other standards can be accessed at the following URL:
http://standards.ieee.org/findstds/errata/index.html. Users are encouraged to check this URL for errata
periodically.
iv
Copyright © 2013 IEEE. All rights reserved.
Patents
Attention is called to the possibility that implementation of this guide may require use of subject matter
covered by patent rights. By publication of this guide, no position is taken by the IEEE with respect to the
existence or validity of any patent rights in connection therewith. If a patent holder or patent applicant has
filed a statement of assurance via an Accepted Letter of Assurance, then the statement is listed on the
IEEE-SA Website http://standards.ieee.org/about/sasb/patcom/patents.html. Letters of Assurance may
indicate whether the Submitter is willing or unwilling to grant licenses under patent rights without
compensation or under reasonable rates, with reasonable terms and conditions that are demonstrably free of
any unfair discrimination to applicants desiring to obtain such licenses.
Essential Patent Claims may exist for which a Letter of Assurance has not been received. The IEEE is not
responsible for identifying Essential Patent Claims for which a license may be required, for conducting
inquiries into the legal validity or scope of Patents Claims, or determining whether any licensing terms or
conditions provided in connection with submission of a Letter of Assurance, if any, or in any licensing
agreements are reasonable or non-discriminatory. Users of this guide are expressly advised that
determination of the validity of any patent rights, and the risk of infringement of such rights, is entirely
their own responsibility. Further information may be obtained from the IEEE Standards Association.
v
Participants
At the time this guide was published, Accredited Standards Committee C63®—Electromagnetic
Compatibility had the following membership:
Daniel Hoolihan, Chair

Jerry Ramie, Secretary
Organization Represented Name of Representative
Alcatel–Lucent Technologies ............................................................................................................... Dheena Moongilan
Alliance for Telecommunications Industry Solutions (ATIS).........................................................................Mel Frerking
.............................................................................................................................................................. James Turner (Alt.)
American Council of Independent Laboratories (ACIL) ................................................................................ Harry Hodes
............................................................................................................................................................... John Repella (Alt.)
American Radio Relay League (ARRL) .................................................................................................... Edward F. Hare
........................................................................................................................................................... Kermit Carlson (Alt.)
Apple, Inc. ..................................................................................................................................................... Fraidun Akhi
.............................................................................................................................................................. Indrandil Sen (Alt.)
AT&T ........................................................................................................................................................ David Chapman
Bureau Veritas ......................................................................................................................................... Jonathan Stewart
........................................................................................................................................................... Mairal Hussain (Alt.)
Cisco Systems ................................................................................................................................................. Andy Griffin
................................................................................................................................................................. David Case (Alt.)
Dell Inc. ..................................................................................................................................................... Richard Worley
Electric Power Research Institute ....................................................................................................................Phil Keebler
ETS-Lindgren .................................................................................................................................................. Zhong Chen
........................................................................................................................................................ Michael Foegelle (Alt.)
Federal Communications Commission (FCC) .............................................................................................. William Hurst
Food and Drug Administration (FDA) ................................................................................................ Jeffrey L. Silberberg
................................................................................................................................................................ Don Witters (Alt.)
Information Technology Industry Council (ITIC) .......................................................................................... John Hirvela
....................................................................................................................................................... Joshua Rosenberg (Alt.)
Institute of Electrical and Electronics Engineers, Inc. (IEEE) ............................................................. Donald N. Heirman
IEEE-Electromagnetic Compatibility Society (EMCS) ......................................................................... H. Stephen Berger
........................................................................................................................................................ Donald Sweeney (Alt.)
Motorola Mobility.......................................................................................................................................... Tom Knipple
Motorola Solutions ...................................................................................................................................... William Elliott
........................................................................................................................................................ Deanna Zakharia (Alt.)
National Institute of Standards and Technology (NIST) .............................................................................. Dennis Camell
Polycom .......................................................................................................................................................... Jeff Rodman
............................................................................................................................................................ Tony Griffiths (Alt.)
Research in Motion (RIM) ............................................................................................................................. Masud Attayi
.............................................................................................................................................................. Daoud Attayi (Alt.)
Samsung Telecommunications .......................................................................................................................Tony Riveria
............................................................................................................................................................. Kendra Green (Alt.)
Society of Automotive Engineers (SAE) ..................................................................................................... Poul Andersen
............................................................................................................................................................... Gary Fenical (Alt.)
Sony Ericsson Mobile Communications ....................................................................................................... Mats Hansson
.................................................................................................................................................................. Jon Kenny (Alt.)
Telecommunication Certification Body (TCB) Council ........................................................................................Art Wall
.......................................................................................................................................................... William Stumpf (Alt.)
Telecommunications Industry Association (TIA) ................................................................................... Stephen Whitesell
TUV-America, Inc. .......................................................................................................................................... Chip Fleury
........................................................................................................................................................... David Schaefer (Alt.)
UL LLC ......................................................................................................................................................... Robert Delisi
............................................................................................................................................................. Jeffrey Moser (Alt.)
U.S. Department of Defense—Joint Spectrum Center ............................................................................. Marcus Shellman
....................................................................................................................................................Michael Duncanson (Alt.)
U.S. Department of the Navy—SPAWAR ....................................................................................................... Chris Dilay
...............................................................................................................................................................David Hilton (Alt.)
vi
Individual Members .................................................................................................................................. Daniel Hoolihan
.........................................................................................................................................................................John Lichtig
.................................................................................................................................................................. Werner Schaefer
............................................................................................................................................................... Ralph M. Showers
.............................................................................................................................................................. David Zimmerman
Members Emeritus ................................................................................................................................ Warren Kesselman
..................................................................................................................................................................... Herbert Mertel
.......................................................................................................................................................... H. R. (Bob) Hofmann
At the time this guide was completed, ASC C63® Subcommittee 1 had the following membership:
Dennis Camell, Chair

Harry Hodes, Vice Chair
Janet O’Neil, Secretary
Mark Arthurs H. R. (Bob) Hofmann Nate Potts

Masud Attayi Daniel Hoolihan Werner Schaefer
Zhong Chen Greg Kiemel Ralph M. Showers
Bob DeLisi Victor Kuczynski Dan Sigoiun
Dean Ghizzone Rick Lombardi Jeffrey L. Silberberg
Andy Griffin Randy Long William Stumpf
Tim Harrington Doug Parker Art Wall
Donald Heirman Ghery Pettit Stephen Whitesell
At the time this guide was completed, the ASC C63® Subcommittee 1 Measurement Uncertainty Working
Group had the following membership:
Bob DeLisi, Chair

Dennis Camell, Vice Chair
Tim Harrington Victor Kuczynski Nate Potts

Donald Heirman Doug Parker Werner Schaefer
Daniel Hoolihan Michael J. Windler
vii
Introduction
This introduction is not part of ANSI C63.23-2012, American National Standard Guide for Electromagnetic
Compatibility—Computations and Treatment of Measurement Uncertainty.
ANSI C63.23 is intended to provide measurement laboratories with guidelines and generally accepted
laboratory practices in the determination of EMI measurement uncertainties. The primary application of
this edition of ANSI C63.23 is for use with ANSI C63.4.
This guide may apply to other C63® standards as appropriate.
viii
Contents
1. Overview .................................................................................................................................................... 1
1.1 Scope ................................................................................................................................................... 1
1.2 Purpose ................................................................................................................................................ 2
2. Normative references.................................................................................................................................. 2
3. Definitions .................................................................................................................................................. 3
4. Basic concepts ............................................................................................................................................ 5

4.1 Introduction to measurement uncertainty ............................................................................................ 5
4.2 Concepts of uncertainty ....................................................................................................................... 5
4.3 Application of uncertainty values in conformity assessment ............................................................... 7
5. EMC measurement uncertainty calculations .............................................................................................. 8

5.1 General ................................................................................................................................................ 8
5.2 Uncertainty contributors common to all conducted and radiated emission EMC measurements—
Type B evaluation ................................................................................................................................ 9
5.3 Uncertainty contributors for conducted emission measurements at a mains port using an LISN—
Type B evaluation .............................................................................................................................. 12
5.4 Uncertainty contributors for radiated emission measurements in the frequency range 30 MHz to
1000 MHz on an open-area test site (OATS) or in a semianechoic chamber (SAC)—Type B
evaluation ........................................................................................................................................... 13
5.5 Uncertainty contributors for radiated emission measurements in the frequency range above 1 GHz—
Type B evaluation .............................................................................................................................. 17
5.6 Type A evaluations of contributors ................................................................................................... 19
5.7 Example spreadsheets ........................................................................................................................ 26
Annex A (informative) Aspects of measurement uncertainty theory ........................................................... 29
Annex B (informative) Bibliography............................................................................................................ 41
ix
Uncertainty
IMPORTANT NOTICE: IEEE Standards documents are not intended to ensure safety, health, or
environmental protection, or ensure against interference with or from other devices or networks.
Implementers of IEEE Standards documents are responsible for determining and complying with all
appropriate safety, security, environmental, health, and interference protection practices and all
applicable laws and regulations.
This IEEE document is made available for use subject to important notices and legal disclaimers.
These notices and disclaimers appear in all publications containing this document and may
be found under the heading “Important Notice” or “Important Notices and Disclaimers
Concerning IEEE Documents.” They can also be obtained on request from IEEE or viewed at
http://standards.ieee.org/IPR/disclaimers.html.
1. Overview
1.1 Scope
In standardized emissions compliance measurements, the emission level of an equipment under test (EUT)
is measured, after which compliance with the associated limit is determined. The measured emission level
is an approximation of the true level due to uncertainties induced by various contributors. In emissions
compliance testing, major relevant contributors related to the EUT or to the measurement procedure happen
to be unspecified and no quantitative information is available about their values, for example, operator
skills, quality system of the laboratory performing the test, environmental conditions, production tolerance
of the EUT model, operation of the EUT, and so on. This document concentrates on the measurement
instrumentation uncertainty, which is a subpart of the total uncertainty of the measurement, and it includes
only the effects of those contributors that are related to the measurement instrumentation.
This application guide provides methods for determining the uncertainty of measurement for
electromagnetic interference (EMI) measurement results. This guide provides information on the
application of Type A statistical evaluations. For Type B applications, this guide also provides information
on where to obtain specified published information that can lead to an evaluation of uncertainty. The
1
Copyright © 2013 IEEE. All rights reserved
ANSI C63.23-2012
American National Standard Guide for Electromagnetic Compatibility—Computations and Treatment of Measurement
Uncertainty
current document provides information on the range 150 kHz to 30 MHz for conducted emissions on main
lines, and 30 MHz to 18 000 MHz for radiated emissions measurements.
1.2 Purpose
This document provides recommended methods for estimating the uncertainty of EMI measurements as
performed in accordance with C63® standards (e.g., ANSI C63.41), for the purposes of this edition, or
similar standards that do not already contain recommended methods for estimating measurement
uncertainty.
2. Normative references
The following referenced documents are indispensable for the application of this document (i.e., they must
be understood and used, so each referenced document is cited in text and its relationship to this document is
explained). For dated references, only the edition cited applies. For undated references, the latest edition of
the referenced document (including any amendments or corrigenda) applies.
ANSI C63.2, American National Standard for Electromagnetic Noise and Field Strength Instrumentation,
10 Hz to 40 GHz Specifications.2
ANSI C63.4, American National Standard for Methods of Measurement of Radio-Noise Emissions from
Low-Voltage Electrical and Electronic Equipment in the Range of 9 kHz to 40 GHz.
ANSI C63.14, American National Standard Dictionary of Electromagnetic Compatibility (EMC) including
Electromagnetic Environmental Effects (E3).
CISPR 16-1-1:2007, Specification for Radio Disturbance and Immunity Measuring Apparatus and
Methods—Part 1-1: Radio Disturbance and Immunity Measuring Apparatus—Measuring Apparatus.3
Methods—Part 1-2: Radio Disturbance and Immunity Measuring Apparatus—Ancillary Equipment—
Conducted Disturbances.
Methods—Part 1-4: Radio Disturbance and Immunity Measuring Apparatus—Antennas and Test Sites for
Radiated Disturbance Measurements—Edition 3.0.
CISPR 16-4-2, Specification for Radio Disturbance and Immunity Measuring Apparatus and Methods—
Part 4-2: Uncertainties, Statistics and Limit Modeling—Uncertainty in EMC Measurements.
ISO 21748:2004, Guidance for the Use of Repeatability, Reproducibility and Trueness Estimates in
Measurement Uncertainty Estimation.4
ISO 21749:2005, Measurement and Uncertainty for Metrological Applications—Repeated Measurements

and Nested Experiments.
1
Information on references can be found in Clause 2.
2
ANSI publications are available from the American National Standards Institute (http://www.ansi.org/).
3
CISPR documents are available from the International Electrotechnical Commission (http://www.iec.ch/). They are also available in
the United States from the American National Standards Institute (http://www.ansi.org/).
4
ISO publications are available from the ISO Central Secretariat (http://www.iso.org/). ISO publications are also available in the
United States from the American National Standards Institute (http://www.ansi.org/).
2
ANSI C63.23-2012
Uncertainty
3. Definitions
For the purposes of this document, the following terms and definitions apply. ANSI C63.14 and the IEEE
Standards Dictionary Online should be consulted for terms not defined in this clause.5 General terms and
definitions used in the expression of uncertainty are provided in ISO/IEC Guide 98-3:2008 [B16]6 (GUM),
and important general terms and definitions related to measurement are contained in ISO/IEC Guide
99:2007 [B17] (VIM).
analysis of variance (ANOVA): A collection of statistical models and their associated procedures in
which the observed variance is partitioned into components due to different explanatory variables. ANOVA
is a general technique that can be used to test the hypothesis that the means among two or more groups are
equal, under the assumption that the sampled populations are normally distributed.
a priori distribution: A distribution in which the probability curve is an isosceles triangle symmetrical
about the center of the base [e.g., the result of casting (rolling) a pair of fair dice]. Syn.: triangular
distribution.
combined standard uncertainty, uc(y): Standard uncertainty of the result of a measurement when that
result is obtained from the values of a number of other quantities, equal to the positive square root of a sum
of terms, the terms being the variance or covariance of these other quantities weighted according to how the
measurement result varies with changes in these quantities.
conformity assessment: Demonstration that specified requirements relating to a product, process, system,
person, or body are fulfilled. [ISO/IEC 17000:2004 [B15], 2.1 and 0.5, MOD]
NOTE 1—The subject field of conformity assessment includes activities such as testing, inspection, and certification,
as well as the accreditation of conformity assessment bodies.7
NOTE 2—The expression “object of conformity assessment” or “object” is used in ISO/IEC 17000:2004 [B15] to
encompass any particular material, product, installation, process, system, person, or body to which conformity
assessment is applied.
NOTE 3—ISO/IEC 17000:2004 [B15] does not include a definition of “conformity” on the grounds that it is not
necessary to do so. “Conformity” does not feature in the definition of “conformity assessment.” The concept of
“conformity assessment” is concerned with “fulfillment of specified requirements,” not with the wider concept of
“conformity.” A definition of the term “specified requirement” is included in ISO/IEC 17000-2004 [B15]. In English,
the term “compliance” is used to distinguish the action of doing what is required (e.g., an organization “complies” by
making something conform or by fulfilling a regulatory requirement).
correction factor: Numerical factor by which the uncorrected result of a measurement is mathematically
adjusted to compensate for an assumed systematic error (e.g., in EMC measurements), the loss in a signal
cable that is usually measured and stated in decibels and added to an uncorrected result that is also
measured and stated in decibels.
NOTE—Because the systematic error cannot be known perfectly, the compensation cannot be complete.
coverage factor, k: Numerical factor used as a multiplier of the combined standard uncertainty in order to
obtain an expanded uncertainty. When k = 2, the confidence level approximates 95% (in case of normal,
Gaussian, probability distribution).
5
The IEEE Standards Dictionary Online subscription is available at http://www.ieee.org/portal/innovate/products/standard/
standards_dictionary.html.
6
The numbers in brackets correspond to those of the bibliography in Annex B.
7
Notes in text, tables, and figures of a standard are given for information only and do not contain requirements needed to implement
this standard.
3
ANSI C63.23-2012
Uncertainty
NOTE—The coverage factor k is typically in the range of 2 to 3, but different values may be used depending on the
desired confidence level and on the type of probability distribution of the combined standard uncertainty.
degrees of freedom—νi,νeff: The degrees of freedom νi of the standard uncertainties based on Type B
evaluation are, by definition, infinite. If the uncertainty is obtained from a limited number of data points or
observations, as in the case of a Type A evaluation, then νi = n – 1 (see ANSI/NCSL Z540-1-1994 [B3],
CISPR/TR 16-4-1:2009 [B7], UKAS LAB 34:2002 [B21], UKAS M 3003:2007 [B22], and ISO/IEC Guide
98-3:2008 [B16]). The effective degrees of freedom νeff are the “combined” degrees of freedom of the
combined standard uncertainty when it is made up of a combination of Type B and Type A evaluations.
drift: The unwanted change of the value of a measurable characteristic of a device over a specified period
of time when the values of all other characteristics of the device are kept constant. For example, drift may
be the change in the device response with use or age, or since the last calibration or measurement. This may
take the form of a change in readout, usually gradual, without associated change in any influence quantity.
error (of measurement): Result of a measurement minus the true value of the measurand.
NOTE 1—The quantity is sometimes called absolute error of measurement when it is necessary to distinguish it from
relative error.
NOTE 2—Because a true value cannot be determined, in practice a conventional true value is used, if available.
expanded uncertainty, U: Quantity defining the interval about the result of a measurement within which
the values that could reasonably be attributed to the measurand may be expected to lie with a high-specified
level of confidence. Expanded uncertainty is obtained by multiplying the combined standard uncertainty by
a coverage factor k.
influence quantity: Quantity that is not necessarily included in the specification of the measurand but that
nonetheless affects the result of the measurement (e.g., frequency in the measurement of an alternating
electric potential difference) or temperature of a micrometer used to measure length.
measurand: Specific quantity subject to measurement.
measurement uncertainty: Non-negative parameter characterizing the dispersion of the quantity values
being attributed to a measurand based on the information used. Syn.: uncertainty of measurement;
uncertainty.
NOTE 1—Measurement uncertainty includes components arising from systematic effects, such as components
associated with corrections and the assigned quantity values of measurement standards, as well as the definitional
uncertainty. Sometimes estimated systematic effects are not corrected for, but instead, associated measurement
uncertainty components are incorporated.
NOTE 2—The parameter may be, for example, a standard deviation called standard measurement uncertainty (or a
specified multiple of it), or the half-width of an interval, having a stated coverage probability.
NOTE 3—Measurement uncertainty comprises, in general, many components. Some of these may be evaluated by
Type A evaluation of measurement uncertainty from the statistical distribution of the quantity values from series of
measurements and can be characterized by standard deviations. The other components, which may be evaluated by
Type B evaluation of measurement uncertainty, can also be characterized by standard deviations, evaluated from
probability density functions based on experience or other information.
NOTE 4—In general, for a given set of information, it is understood that the measurement uncertainty is associated
with a stated quantity value attributed to the measurand. A modification of this value results in a modification of the
associated uncertainty (for example, different measurement uncertainty values for a radiated emissions measurement
result may apply in different frequency ranges).
4
ANSI C63.23-2012
Uncertainty
NOTE 5—This definition of measurement uncertainty (i.e., as repeated from 2.26 of ISO/IEC Guide 99:2007 [B17]),
and for situations where uncertainty components arising from sampling are not considered, corresponds to the term and
definition for standards compliance uncertainty given in 3.1.16 of CISPR 16-4-1:2009 [B7].
4. Basic concepts
4.1 Introduction to measurement uncertainty
Measurement uncertainty is the best estimated quantity by which a measured value differs from the true
value of a parameter under evaluation. Correction factors (or biases; see the definition in A.1.4) are
typically used to improve correlation between measurement devices and systems and the reference quantity
to which they have been calibrated. Additionally, a correction factor is always accompanied by a
measurement uncertainty. The need for such factors is an indication that the true value cannot be directly
and completely obtained from the instrumentation and with the method of measurement used. Correction
factors determined during equipment calibration processes determine the bias to be applied to the
measurement result, and they are subject to measurement uncertainty.
Determining the operational characteristics of the EUT with respect to electromagnetic compatibility
compliance requires both emission and immunity measurements. This guide, however, only focuses on
emission measurement, which is covered in, for example, ANSI C63.4. The measurements or evaluations
involve the use of various instrumentation and techniques, requiring operator interaction and decisions. The
measurement (or evaluation) process produces both random and systematic effects that influence the
ultimate outcome of the measurement. Examples of systematic effects are bias of the measurement result
by cable loss, parallax when reading a D’Arsonval meter, or habits and preferences of the operator. An
example of a random effect is the influence of noise contribution to the measured amplitude if the
amplitude of the measurand is close to the system noise floor. Additional papers on the treatment of
uncertainty components (influence quantities) are listed in an informative bibliography (i.e., see Annex B).
An evaluation of each effect included in a measurement enables the identification of influence factors (i.e.,
elements that contribute to the uncertainty of the measurement), and a thorough description of each effect
can assist in assignment of an appropriate value and weighting factor indicating how the overall
measurement uncertainty budget may be affected. Each element of the measurement process involves an
uncertainty of measurement that can be assigned a value and described by the owner of the process. The
combinations of the elements of uncertainty form the basis of an uncertainty budget for a particular
measurement process.
4.2 Concepts of uncertainty
4.2.1 General
Measurement uncertainty is a parameter associated with the result of a measurement that characterizes the
dispersion of the values that could reasonably be attributed to the true value of the measurand. It shows the
spread of values above and below the measurement result within which the true value of the measurand
may be expected to lie. It can be considered to be a measure of the possible error in the value of the
measurand provided by the result of a measurement; however, the true value of a measurand can never be
known. Measurement uncertainty should not be confused with correction or correlation factors. Correction
and correlation factors are quantities added to measurement values to create agreement between established
standard values and measurement device readings (these can also be described as biases: see A.1.4).
5
ANSI C63.23-2012
Uncertainty
Two types of evaluations are defined to determine the values for an uncertainty budget: Type A and Type
B. The type of evaluation used depends on what input and reference data is available. Other
electromagnetic compatibility (EMC) measurement uncertainty documents, such as CISPR 16-4-2 (an
international standard) and UKAS LAB 34:2002 [B21] (a UKAS guideline document), describe primarily
the Type B analyses and may use Type A methods to obtain only a few contributors, whereas Clause 5 of
this guide describes how to apply either method.
4.2.2 Type A evaluation
Type A evaluations of uncertainty are those obtained by using statistical methods where multiple
observations of the same event are recorded. These observed values are used to calculate the standard
deviation of the results. The standard deviation is then used to obtain the contribution of the process under
observation to the uncertainty budget.
4.2.3 Type B evaluation
Type B evaluations of uncertainty are those by any method other than a statistical evaluation. Knowledge
of previous performance of instruments, specifications, instrumentation reference data, and uncertainty data
provided with calibrations are examples of this evaluation type.
4.2.4 Type A or Type B evaluation
Either a Type A or a Type B analysis shall be carried out for each contributor to the uncertainty budget. A
Type A analysis has the advantage of providing a more representative evaluation because it is based on the
actual test setup/equipment used during measurement and, as such, generally results in a smaller
contributor value than a Type B analysis. However, a Type A analysis is more demanding in terms of time
and resources because it usually involves testing or analyzing extensive amounts of data.
A Type A evaluation of a contributor is determined by following the guidance for that particular
contributor (see 5.6) and finding the standard deviation for the data collected. For example, antenna factor
accuracy (also called antenna factor calibration) can be determined by taking the calibration value at each
frequency over the past five calibrations and determining the worst-case standard deviation over the
antenna’s usable frequency range. This could result in a smaller contributor than the Type B contributor. If
the Type B contributor for antenna factor accuracy was 0.6 dB, and the Type A analysis of the data yielded
a contributor of 0.2 dB, then it would be advantageous to use the smaller contributor.
4.2.5 Measurement uncertainty budget calculation
Annex A includes a detailed presentation of measurement uncertainty concepts and methods of calculation
of contributors’ contribution to the measurement uncertainty budget based on their known or estimated
probability distribution.
Table 1 summarizes the formulas to be used for each type of probability distribution and provides cross-
references to corresponding subclauses from Annex A.
6
ANSI C63.23-2012
Uncertainty
Table 1 —Calculation of contributors to the measurement uncertainty budget

Evaluation Probability Divisor, k Contributor, uS Annex A
type distribution subclause
Type A Normal (Gaussian) 1 Experimental standard A.2.1.1
deviation of the mean,
s (q )
Type A t-distribution 1 Adjusted experimental A.2.1.2
standard deviation of the
mean, η (ν ) s( q )
Type B Normal (Gaussian) Depends on expanded uncertainty A.2.2.2
confidence level
(usually k = 1 or 2; k
see Table A.1)
Type B t-distribution kp (Table A.3) expanded uncertainty A.2.2.3
k
p
Type B Rectangular A.2.2.4
3 a 3
Type B U-shaped A.2.2.5
2 a 2
Type B Triangular A.2.2.6
6 a 6
NOTE 1— s (q ) is the standard deviation of the mean, q , of the population of repeated observations, q .
j
()
NOTE 2— η ν is an expansion coefficient, used for t-distributed Type A analysis, which depends on the
number of degrees of freedom, ν . See Table A.2.
NOTE 3—kp is the coverage factor for the level of confidence p. For the t-distribution, see Table A.3.
NOTE 4—a is the half-width of the probability distribution, a = (a+ − a-) / 2, where a+ and a– define the
interval given by the prescribed limits of the individual uncertainty contributor.
4.3 Application of uncertainty values in conformity assessment
Several approaches are available to the application of measurement uncertainty when determining
compliance of a EUT with EMC standards (e.g., limits). Some documents, like CISPR 16-4-2, include a
process for the application of calculated measurement uncertainty to the process of determining compliance
with emission limits; this procedure is used in documents like CISPR 11-2010 [B10]. Other documents,
like CISPR 22-2010 [B11], state that uncertainty must be calculated and reported but not used in adjusting
measured emission values. Other guidance documents have also been used by laboratories and
accreditation bodies. For example, Figure 1 in UKAS LAB 34:2002 [B21] shows various situations for
evaluating whether measurement results comply with a limit. Please refer to UKAS LAB 34:2002 [B21] for
more details on the application of measurement uncertainty to limits or measurements.
This approach is not considered herein because this document is intended to address only measurement
instrumentation uncertainty and associated parameters within the measurement chain. Adding uncertainty
to limits (other than the uncertainty of the measurement instrumentation chain) would need to be based on
either the interference potential of the device measured or the immunity characteristics of given products.
Such interference and immunity analyses are beyond the scope of this document.
7
ANSI C63.23-2012
Uncertainty
5. EMC measurement uncertainty calculations
5.1 General
This clause describes uncertainty contributors for the conducted and radiated emission EMC test methods,
and it illustrates EMC measurement instrumentation uncertainty estimations using example numeric values.
The contributors described in this clause are based on measurements performed as per ANSI C63.4. In the
case of other measurement standards, the list of contributors shall be adjusted accordingly to reflect the
measurement method therein.
Each method is divided into two evaluations, as follows:
⎯ Descriptions of contributors and equations used to calculate standard, combined, and expanded
uncertainties, and example Type B uncertainty template
⎯ Type A uncertainty statistical estimation
The Type B uncertainty for each test method is addressed in 5.2 through 5.5. The Type A uncertainties
apply to all methods and are addressed in 5.6. Example spreadsheets are included in 5.7.
Numerical examples are included for some contributors in the following subclauses. These are given only
for illustration purposes, and they are not meant to be used indiscriminately when preparing the
measurement uncertainty budget, as they may not be appropriate for a particular test laboratory. Each
laboratory shall consider the information available about its own particular measuring system, including
equipment characteristics, actual validation data of the test sites, the specific calibration data, the known or
likely probability distributions, and internal measurement procedures.
Table 2 lists all contributors to the measurement uncertainty budget, for both conducted and radiated
emissions test methods.
8
ANSI C63.23-2012
Uncertainty
Table 2 —List of contributing factors

Contributor Test method Subclause Subclause
(Type A (Type B
evaluation) evaluation)
Receiver reading Conducted and radiated 5.6.2.1 5.2.1
emissions
Attenuation of the connection between signal Conducted and radiated 5.6.2.2 5.2.2
transducer and receiver emissions
Receiver sine-wave voltage accuracy Conducted and radiated 5.6.2.3 5.2.3
emissions
Receiver pulse amplitude response Conducted and radiated 5.6.2.4 5.2.4
emissions
Receiver pulse response variation with repetition Conducted and radiated 5.6.2.5 5.2.5
frequency emissions
Receiver noise floor proximity Conducted and radiated 5.6.2.6 5.2.6
emissions
Mismatch effects between the receiver port of the Conducted and radiated 5.6.2.7 5.2.7
signal transducer and receiver emissions
Preamplifier gain variation Conducted and radiated 5.6.2.8 5.2.8
emissions
Instability of preamplifier Conducted and radiated 5.6.2.9 5.2.9
emissions
Line impedance stabilization network (LISN) Conducted emissions 5.6.3.2 5.3.2
voltage division factor
LISN impedance Conducted emissions 5.6.3.3 5.3.3
Antenna factor accuracy Radiated emissions 5.6.4.2 5.4.2 and 5.5.2
Antenna factor frequency interpolation Radiated emissions 5.6.4.3 5.4.3 and 5.5.3
Antenna factor variation with height Radiated emissions 5.6.4.4 5.4.4
Antenna directivity Radiated emissions 5.6.4.5 5.4.5 and 5.5.4
Antenna phase center location Radiated emissions 5.6.4.6 5.4.6 and 5.5.5
Antenna cross-polarization response Radiated emissions 5.6.4.7 5.4.7 and 5.5.6
Antenna balance Radiated emissions 5.6.4.8 5.4.8
Test site contribution Radiated emissions 5.6.4.9 5.4.9 and 5.5.7
Separation between equipment under test and Radiated emissions 5.6.4.10 5.4.10 and
measurement antenna 5.5.8
Height of table supporting the equipment under test Radiated emissions 5.6.4.11 5.4.11
Effect of setup table material supporting the Radiated emissions 5.6.4.12 5.4.12 and
equipment under test 5.5.9
5.2 Uncertainty contributors common to all conducted and radiated emission EMC
measurements—Type B evaluation
5.2.1 Receiver reading
The receiver reading will vary for different reasons, such as input attenuation switching uncertainty,
detector nonlinearity, and marker resolution (round-off error). Estimates of these contributors can be taken
from manufacturer specifications or from the calibration certificate of the receiver. A rectangular
probability distribution shall be used in either case. The use of calibration certificate data will significantly
reduce the values of these contributors; however, in this case, the uncertainties of measurement for each
contributor, as reported by the calibration laboratory, must be included in the overall uncertainty budget.
9
ANSI C63.23-2012
Uncertainty
EXAMPLE 1—The calibration certificate reports a worst-case deviation from linearity (detector nonlinearity) of
+0.12 dB/−0.11 dB. Using a rectangular distribution, the contributor value to be used in the overall budget will be [0.12
− (−0.11)] / (2 √3) = 0.066 dB. See A.2.2.4 for details on how to calculate the contributor value in case of rectangular
probability distribution.
EXAMPLE 2—The calibration laboratory reported an expanded measurement uncertainty for detector nonlinearity
measurement of ±0.078 dB, also stating that this is based on a normal distribution with a coverage factor k = 2. This
must be included as a separate contributor in the overall budget as [0.078 − (−0.078)] / (2k) = 0.039 dB. See A.2.2.2 for
details on how to calculate the contributor value in case of normal (Gaussian) probability distribution.
When evaluating the contributors affecting the receiver reading, the user must be careful not to double-
count specific contributors. For example, a frequency response will be included in the receiver sine-wave
voltage accuracy, which is separately considered in 5.2.3. For more details, refer to Medler [B18] and
Stecher [B20].
5.2.2 Attenuation of the connection between signal transducer and receiver
An estimate of the attenuation of the connection between the receiver and the signal transducer (LISN,
current probe, or antenna) is assumed to be available from a calibration report, along with an expanded
uncertainty and a coverage factor. If the estimate of attenuation is obtained from manufacturer’s data for a
cable or attenuator, then a rectangular probability distribution having a half-width equal to the
manufacturer’s specified tolerance on the attenuation may be assumed. If the connection is a cable and an
attenuator, with the manufacturer’s data available on each, then this contribution has two components, each
with its own rectangular probability distribution.
5.2.3 Receiver sine-wave voltage accuracy
An estimate of the magnitude for receiver sine-wave voltage accuracy is assumed to be available from a
calibration report, along with an expanded uncertainty and a coverage factor. If a calibration report states
only that the receiver sine-wave voltage accuracy meets the ANSI C63.2 specification (i.e., ±2 dB), the
estimate of the contributing factor should be taken as zero with a rectangular probability distribution having
a half-width of 2 dB.
5.2.4 Receiver pulse amplitude response
In general, it is not practical to correct for receiver pulse response characteristics. A verification report
stating that the receiver pulse amplitude response complies with the CISPR 16-1-1:2007 tolerance of
±1.5 dB (at the reference pulse repetition frequency) for peak, quasi-peak, or average detection is assumed
to be available. The magnitude is estimated to be zero with a rectangular probability distribution having a
half-width of 1.5 dB.
If an emission is narrowband in nature [e.g., continuous wave (CW) signal], then pulse response corrections
need not be considered.
5.2.5 Receiver pulse response variation with repetition frequency
In general, it is not practical to correct for receiver pulse response characteristics. The CISPR 16-1-1
tolerance for pulse repetition rate response varies with repetition rate and detector type. A verification
report stating that the receiver pulse repetition rate responses comply with the CISPR 16-1-1 tolerances is
assumed to be available. The correction is estimated to be zero with a rectangular probability distribution
having a half-width of 1.5 dB, a value considered to be representative of the various CISPR 16-1-1
tolerances.
10
ANSI C63.23-2012
Uncertainty
If an emission is narrowband in nature (e.g., CW signal), then pulse response corrections need not be
considered.
5.2.6 Receiver noise floor proximity
The noise floor of an EMI receiver meeting CISPR requirements is usually sufficiently below the emission
voltage or current limit that its effect is negligible on measurement results close to these limits. However,
for radiated emissions measurements, the proximity of the receiver noise floor may influence measurement
results near the radiated emissions limit. For radiated emissions measurements, the magnitude is estimated
to be zero with an expanded uncertainty of 0.5 dB and a coverage factor of 2 (normal distribution).
5.2.7 Mismatch effects between the receiver port of the signal transducer and the receiver
In general, the receiver port of a signal transducer (LISN, current probe, or antenna) will be connected to
port 1 of a two-port network whose port 2 is terminated by a receiver of reflection coefficient Γr. The two-
port network, which might be a cable, an attenuator, a combination of attenuator and cable, or some other
combination of components, can be represented by its S-parameters. The mismatch correction is then (see
UKAS LAB 34:2002 [B21], Carpenter [B5], and Carpenter [B6]):
[
δM = 20 log10 ( 1 − Γe S11 )( 1 − Γr S 22 ) − S 21
2
Γe Γr ] (1)
where Γe is the reflection coefficient observed looking into the receiver port of the LISN with the EUT
connected, or looking into the output port of the antenna when it is set up for disturbance measurement. All
parameters are with respect to 50 Ω.
When only the magnitudes, or extremes of magnitudes, of the parameters are known, it is not possible to
calculate δM, but its extreme values δM± are not greater than:
[ (
δM ± = 20 log10 1 ± Γe S11 + Γr S 22 + Γe Γr S11 S 22 + Γe Γr S 21
2
)] (2)
The probability distribution of δM is approximately U-shaped, with width not greater than (δM+ − δM−) and
standard deviation not greater than the half-width divided by √2.
It is usually assumed that the connection to the receiver from the transducer is a well-matched cable (i.e.,
S11 << 1 and S22 << 1 ) of negligible attenuation (i.e., S 21 ≈ 1 ), and that the receiver radio frequency (RF)
attenuation is set to 10 dB or more, for which the CISPR 16-1-1:2007 voltage standing wave ratio (VSWR)
tolerance of VSWR ≤ 1.2:1 can be assumed for the calculation, as a worst-case estimate.
In radiated emission measurements, the VSWR of the antenna can be a major contributor to the
measurement uncertainty, due to the resulting mismatch effect. The VSWR of antenna used shall be
known, and well-matched attenuators should be used at the antenna connector, as necessary, to meet
antenna VSWR requirements per CISPR 16-1-4:2010.
If the receiver RF attenuation is set to 0 dB, then the corresponding VSWR must be used for mismatch
calculations. Likewise, if the receiver input attenuation is set to 10 dB or more, then the associated lower
VSWR is used. The VSWR specifications per CISPR 16-1-1:2010 are shown in Table 3.
11
ANSI C63.23-2012
Uncertainty
Table 3 —Receiver input VSWR specifications per CISPR 16-1-1:2010

Frequency range Input attenuation Maximum VSWR
0 dB 2.0:1
9 kHz to 1 GHz
10 dB or higher 1.2:1
0 dB 3.0:1
1 GHz to 18 GHz
10 dB or higher 2.0:1
If a preamplifier external to the receiver is used, then its contribution to the mismatch uncertainty due to the
input and output VSWR is also to be included in the overall uncertainty calculation. In this case, two
mismatch uncertainties have to be considered: between antenna/LISN and preamplifier input port and
between the preamplifier output port and the receiver input port. Please note that the values used for these
calculations must be supported by calibration of the instruments and accessories.
5.2.8 Preamplifier gain variation
Calibrated preamplifiers are used either at the input of the measuring receiver or built into the measuring
receiver. Usually, external preamplifiers are only used for radiated emissions above 1 GHz; however, they
may be occasionally also used for radiated emissions within 30 MHz to 1000 MHz or even for conducted
emissions measurements.
If an external preamplifier is used in the test setup, then the accuracy of its gain calibration shall be
considered as an additional uncertainty contribution. It shall be obtained from a calibration certificate,
which has to provide the expanded uncertainty and a coverage factor. The magnitude is estimated to be
0.4 dB (expanded uncertainty) with a coverage factor of 2 (normal distribution).
5.2.9 Instability of preamplifier
Any preamplifier gain variations that are not taken into account by the internal built-in calibration process
of the receiver need to be considered in the uncertainty estimate. An estimate of the preamplifier gain is
assumed to be available from a calibration report, along with an expanded uncertainty and a coverage factor
(see 5.2.8). Any additional gain deviations from the calibrated frequency response (e.g., due to temperature
changes in the test environment or aging) have to be taken into account as an uncertainty contribution. The
estimate of the magnitude for this uncertainty component is ±0.7 dB with a rectangular distribution.
5.3 Uncertainty contributors for conducted emission measurements at a mains

port using an LISN—Type B evaluation
5.3.1 General
In addition to the contributors included in 5.2, uncertainty quantities to be considered for conducted
disturbance measurements at a mains port using an LISN are as follows.
5.3.2 LISN voltage division factor
An estimate of the LISN voltage division factor is assumed to be available from a calibration report, along
with an expanded uncertainty and a coverage factor.
12
ANSI C63.23-2012
Uncertainty
5.3.3 LISN impedance
The impedance tolerance in ANSI C63.4 for a 50 Ω/50 μH + 5 Ω AMN or a 50 Ω/50 μH LISN requires the
impedance to be within 20% of the magnitude of the nominal impedance when the receiver port is
terminated in 50 Ω. The estimate of the correction is zero with a probability distribution bounded by the
extremes from all combinations of the constrained LISN impedance and unconstrained EUT impedance
over the defined frequency range. A triangular probability distribution is assumed because there is a low
probability of encountering the particular combinations of frequency, LISN impedance, and EUT
impedance needed to produce those extremes.
5.4 Uncertainty contributors for radiated emission measurements in the frequency

range 30 MHz to 1000 MHz on an open-area test site (OATS) or in a semianechoic
chamber (SAC)—Type B evaluation
5.4.1 General
In addition to the contributors included in 5.2, uncertainty quantities to be considered for radiated emission
measurements on an OATS or in a SAC, in the frequency range 30 MHz to 1000 MHz, are as follows.
5.4.2 Antenna factor accuracy
An estimate of the free-space antenna factor is assumed to be available from a calibration report, along with
an expanded uncertainty and a coverage factor. The uncertainty of the antenna calibration may vary,
dependent on the antenna type, the applied calibration method, and the frequency range.
5.4.3 Antenna factor frequency interpolation
When an antenna factor value is calculated by linear interpolation between frequencies at which calibration
data are available, the uncertainty associated with that antenna factor depends on the frequency interval
between calibration points and the variability of the antenna factor with frequency. Plotting the calibrated
antenna factor versus frequency can help to estimate the actual magnitude of this contribution. The
uncertainty can be reduced by requesting a large number of frequency points (e.g., 1601) from the
calibration laboratory, which can easily be achieved by using a vector network analyzer for the antenna
calibration process.
5.4.4 Antenna factor variation with height
The antenna factor changes due to mutual coupling of the antenna with its image in the ground plane. When
an antenna is scanned in height above a well-conducting ground plane, the average antenna factor is close
in magnitude to the free-space antenna factor. The height scan must be at least half a wavelength with
readings taken at intervals of one eighth of a wavelength or less, with the lowest height greater than a third
of a wavelength. The change in mutual impedance is most sensitive for tuned dipoles. Assuming that the
longest dipole is tuned at 80 MHz, a height of no more than 4 m is required. The uncertainty contribution is
the possible deviation from the free-space antenna factor. If this uncertainty contribution varies
significantly over the frequency range, then a correction is applied to the measurement data for each
frequency range. In general, this contribution decreases as the frequency increases, and it becomes
negligible above 300 MHz.
13
ANSI C63.23-2012
Uncertainty
5.4.5 Antenna directivity
CISPR 16-1-4:2010 requires the responses of a complex antenna in the direction of the direct ray and in the
direction of the ground-reflected ray to be within 1 dB of the maximum response. To meet this requirement,
the main lobe of the complex antenna may need to be tilted downward, particularly at separations of less
than 10 m. The uncertainty contribution for the effect of directivity is 0 dB for an antenna having a uniform
pattern in the vertical plane, and between 0 dB and +xi dB for an antenna having a nonuniform pattern in
the vertical plane, where xi dB is given in Table 4 through Table 7 for horizontally aligned and for tilted
antennas. CISPR 16-1-4:2010 gives guidance on the maximum gain allowable for a biconical or
logarithmic-periodic antenna for which the values of xi in Table 4 through Table 7 apply.
A horizontally polarized biconical antenna is assumed to have a uniform pattern in the vertical plane. A
vertically polarized biconical antenna, and a horizontally or vertically polarized log-periodic antenna, is
assumed to require a correction of up to +xi/2 dB at separations of 3 m and 10 m, but not more than +0.15
dB at a separation of 30 m.
The estimate of the correction is +xi/2 dB with a rectangular probability distribution of the uncertainty
having half-width of xi/2 dB.
NOTE 1—The reduction in signal strength caused by reduced directivity at angles off antenna bore sight is a
systematic error and therefore can be corrected; a nonzero estimate with reduced uncertainty can be evaluated from the
known pattern of the measuring antenna and applied as a function of frequency and separation.
NOTE 2—In Table 4 through Table 7, the uncertainty contribution is +xi/2 dB for vertical polarization with an EUT
height of 1.5 m. The value is positive because it represents a reduction in field strength.
Table 4 —Horizontally polarized radiated emissions from 30 MHz to 200 MHz

using a biconical antenna at distances of 3 m, 10 m, or 30 m
Magnitude Probability Normalized
Uncertainty contribution
(dB) distribution magnitude (dB)
Directivity difference at 3 m ±0.0 Rectangular 0.00
at 10 m ±0.0 Rectangular 0.00
at 30 m ±0.0 Rectangular 0.00
Table 5 —Vertically polarized radiated emissions from 30 MHz to 200 MHz using a
biconical antenna at distances of 3 m, 10 m, or 30 m
Directivity difference at 3 m (<130 MHz) +1.0/–0.0 Rectangular 0.29
at 3 m (>130 MHz) +2.0/–0.0 Rectangular 0.58
at 3 m with tilting +1.0/–0.0 Rectangular 0.29
at 10 m +0.5/–0.0 Rectangular 0.14
14
ANSI C63.23-2012
Uncertainty
Table 6 —Horizontally polarized radiated disturbances from 200 MHz to 1 GHz

using a logarithmic-periodic antenna at a distance of 3 m, 10 m, or 30 m
Directivity difference at 3 m +2.0/–0.0 Rectangular 0.58
Directivity difference at 3 m with tilting +1.0/–0.0 Rectangular 0.29
Table 7 —Vertically polarized radiated disturbances from 200 MHz to 1 GHz

using a logarithmic-periodic antenna at a distance of 3 m, 10 m, or 30 m
Directivity difference at 3 m +2.5/–0.0 Rectangular 0.73
Directivity difference at 3 m with tilting +1.2/–0.0 Rectangular 0.35
5.4.6 Antenna phase center location
The uncertainty contribution due to phase center location variation is negligible for a biconical antenna, but
the change in phase-center location with frequency for a log-periodic antenna causes a deviation from the
required measurement distance.
For a log-periodic antenna, the magnitude is determined by considering the effect of an error of ±0.35 m in
the separation and assuming that field strength is inversely proportional to separation.
NOTE 1—If a dipole is the measuring antenna, then this uncertainty contribution is negligible.
NOTE 2—For hybrid antennas, the magnitude will increase if no correction for the phase center variation is applied for
this systematic effect.
5.4.7 Antenna cross-polarization response
The cross-polarization response of a biconical antenna is considered to be negligible. The estimate of the
cross-polarization response of a log-periodic antenna is assumed to be ±0.9 dB with a rectangular
distribution, corresponding to the CISPR 16-1-4:2010 cross-polarization response tolerance of ±20 dB.
Cross polarization error is defined as:
 y−x

20 log10 1 ± 10 20  (3)
 
 
where x is the linear response and y is the cross-polarized response in dBµV.
The cross-polarization error of the antenna used for measurements can be measured by placing it in a plane
wave. The antenna under test is set to the same polarization as the transmit antenna (vertical polarization is
better because it minimizes ground reflections), and the value of x is recorded. Then, without changing the
transmitting antenna, the receive antenna is changed to the opposite polarization, and the value of y is
recorded. This test must be performed at a sufficient number of frequencies evenly spread across the
30 MHz to 1000 MHz range. The magnitude of the worst-case cross-polarization error thus measured is
considered as the contributor’s value, with a rectangular distribution.
15
ANSI C63.23-2012
Uncertainty
NOTE 1—At the receive antenna, a plane wave must be realized (no ground or wall reflections).
NOTE 2—If a dipole is used as the measuring antenna, then the cross-polarization response is negligible.
5.4.8 Antenna balance
The effect of an unbalanced antenna is greatest when the input coaxial cable is aligned parallel to the
antenna elements. Antenna balance is predominantly an issue for biconical antennas. It is caused by the
imbalance of the balun used to feed the elements of the biconical antenna. This imbalance causes the
coupling of surface currents on to the cable. The estimate of the magnitude is based on the performance of
commercially available antennas, with a rectangular distribution. Table 8 summarizes the values.
Table 8 —Estimate of antenna balance contributor

Antenna Polarization Magnitude Normalized
magnitude
Biconical Horizontal ±0.3 0.17
Vertical ±0.9 0.52
Logarithmic-periodic Horizontal ±0.0 0.0
Vertical ±0.3 0.17
5.4.9 Test site contribution
The magnitude Δmax of the maximum difference between theoretical site attenuation, on the one hand, and
measured site attenuation increased by the site attenuation measurement uncertainty, on the other hand,
provides an indication of the effect that site imperfection may have on a disturbance measurement. The
tolerance for this difference per ANSI C63.4 is ±4 dB. However, the measurement uncertainty associated
with the site-attenuation measurement method is usually large and is dominated by the two antenna factor
uncertainties. Therefore, a site that meets the 4 dB tolerance is unlikely to have imperfections sufficient to
cause errors of 4 dB in disturbance measurements. In recognition of this fact, a triangular probability
distribution is assumed for the determination of the impact of this uncertainty component. Therefore, the
worst-case estimate for the magnitude of this contribution is ±4 dB with a triangular probability
distribution.
NOTE—If Δmax is less than 4 dB, then the estimate can be assumed as the actual Δmax with a triangular probability
distribution.
5.4.10 Separation between equipment under test and measurement antenna
The error in separation is due to errors in determining the perimeter of the EUT and distance measurement.
The magnitude of this uncertainty contribution is based on assuming a maximum separation error of ±0.1 m
and on the assumption that field strength is inversely proportional to separation over that distance margin.
5.4.11 Height of table supporting the equipment under test
This uncertainty is caused by the placement of the equipment under test on a table of a height other than the
nominal height of 0.8 m. The magnitude estimate is based on a maximum deviation of ±0.01 m from the
nominal table height, resulting in an expanded uncertainty of 0.1 dB. A normal distribution is assumed and
a coverage factor k = 2 is used, for approximately 95% level of confidence.
For floor-standing equipment, this uncertainty contribution does not apply.
16
ANSI C63.23-2012
Uncertainty
5.4.12 Effect of table material supporting the equipment under test
CISPR 16-1-4:2010 describes a method for the evaluation of the effect of the table material up to 1 GHz.
No tolerance is given for this effect. Below 200 MHz, the magnitude of this uncertainty contribution is
assumed to be zero. Above 200 MHz, the estimate of the magnitude is ±0.5 dB with a rectangular
distribution function. The actual magnitude for a specific setup must be derived from measurements, using
the procedure outlined in CISPR 16-1-4:2010.
5.5 Uncertainty contributors for radiated emission measurements in the frequency

range above 1 GHz—Type B evaluation
5.5.1 General
In addition to the contributors included in 5.2, uncertainty quantities to be considered for radiated emission
measurements above 1 GHz in a quasi–free-space environment are as follows.
5.5.2 Antenna factor accuracy
An estimate of the free-space antenna factor is assumed to be available from a calibration report, along with
an expanded uncertainty and a coverage factor. The stated uncertainty for the antenna factor is to be used in
the measurement uncertainty calculation, not the antenna factor itself. The uncertainty of the antenna
calibration may vary, dependent on the antenna type, the applied calibration method, and the frequency
range.
5.5.3 Antenna factor frequency interpolation
When an antenna factor value is calculated by interpolation between frequencies at which calibration data
are available, the uncertainty associated with that antenna factor depends on the frequency interval between
calibration points and the variability of antenna factor with frequency. Plotting calibrated antenna factor
versus frequency may help to estimate the actual magnitude of this contribution. The uncertainty can be
reduced by requesting a large number of frequency points (e.g., 1601) from the calibration laboratory,
which can easily be achieved by using a vector network analyzer for the antenna calibration process.
5.5.4 Antenna directivity
The directivity of the receive antenna determines the value w [i.e., per Equation (10) of
CISPR 16-2-3:2006-2007 [B11], reproduced below], which is used to determine the need for an antenna
height scan. The parameter w is calculated assuming that the far-field criterion is valid. At close
measurement distances, measurements are made in the Fresnel zone and not in the far field. The actual
dimension w is different from the value obtained from Equation (4):
(
w = 2d tan 0.5θ 3 dB ) (4)
where
d is the measurement distance

q3 dB is the receive antenna beam width at the 3 dB points
17
ANSI C63.23-2012
Uncertainty
The impact of the receive antenna properties on the uncertainty is also determined by frequency, the size of
the EUT, and the measurement distance. The resulting value of the uncertainty is not easily determined. At
higher frequencies, some receive antennas have multiple lobes instead of one main lobe. This may cause
additional instrumentation uncertainties, which are not considered here. The magnitude estimate of this
uncertainty parameter is +3/–0 dB with a rectangular probability distribution, assuming that the EUT
dimension is larger than the parameter w, based on the antenna radiation pattern.
5.5.5 Antenna phase center
The uncertainty contribution due to phase center location for a horn antenna causes a deviation from the
required measurement distance.
For a horn antenna, the magnitude is determined by considering the effect of an error of ±0.10 cm in the
separation and assuming that field strength is inversely proportional to separation.
NOTE 1—If a dipole is the measuring antenna, then this uncertainty contribution is negligible.
NOTE 2—For hybrid antennas, the magnitude will increase if no correction for the phase center variation is applied for
this systematic effect.
5.5.6 Antenna cross-polarization response
The cross-polarization response of a dipole type antenna is considered to be negligible. The estimate of the
cross-polarization response of a log-periodic antenna is assumed to be ±0.9 dB with a rectangular
distribution, corresponding to the CISPR 16-1-4:2010 cross-polarization response tolerance of 20 dB. The
cross-polarization response for a horn antenna is considered negligible.
5.5.7 Test site
The measured site voltage standing wave ratio SVSWR provides an indication of the effect that site
imperfection may have on an emission measurement. CISPR 16-1-4:2010 states a tolerance for the SVSWR
of 6 dB. However, a site that meets the 6 dB SVSWR tolerance will not cause errors of 6 dB in disturbance
measurements. A useful comparison between SVSWR and deviation from the reference transmission loss for
a 3 m site has been made in document CISPR/A/838/INF (2009) [B7]. In this document, it is shown that a
SVSWR of 6 dB corresponds roughly to a 4 dB deviation from transmission loss. Assuming that the
transmission loss has a Gaussian probability distribution and because the value of 4 dB is not exceeded in
the entire frequency range, a value of 4 dB is assumed as the corresponding expanded uncertainty with a
coverage factor k = 3 (i.e., the standard uncertainty is 1.33 dB).
NOTE—If SVSWR is less than 6 dB, then the estimate of the uncertainty contribution may be determined as
[4 × (SVSWR / 6)] dB, assuming a normal distribution and a coverage factor k = 3.
5.5.8 Separation between equipment under test and measurement antenna
The error in separation is due to errors in determining the perimeter of the EUT and distance measurement.
The magnitude of this uncertainty contribution is based on assuming a maximum separation error of ±0.1 m
and on the assumption that field strength is inversely proportional to separation over that distance margin.
18
ANSI C63.23-2012
Uncertainty
5.5.9 Effect of table material supporting the equipment under test
CISPR 16-1-4:2010 describes a method for the evaluation of the effect of the table material above 1 GHz.
No tolerance is given for this effect. From 1 GHz to 6 GHz, the estimated magnitude of this uncertainty
contribution is ±1.5 dB with a rectangular distribution function. Above 6 GHz, the estimate of the
magnitude is ±2.0 dB with a rectangular distribution function. The actual magnitude for a specific setup
must be derived from measurements, using the procedure outlined in CISPR 16-1-4:2010.
5.6 Type A evaluations of contributors
5.6.1 General
Type A evaluation of individual contributors can be used to assess uncertainty as long as the evaluation
methodology is sound and based on a statistical analysis of data. The following are examples of the
methods that can be employed to assess a contributor. These examples are by no means the only approach.
The specific contributors listed in this subclause are not the only areas where a Type A evaluation approach
can be used. These are merely examples of how to apply such an approach. Type A evaluations of test
systems, as opposed to contributors, are also possible. For further information on using repeatability and
reproducibility studies including interlaboratory comparisons in Type A studies, see ISO 21748:2004 and
ISO 21749:2005.
When the number of observations made is sufficiently large, a normal distribution can be assumed and the
experimental standard deviation of the mean is used as the uncertainty value. However, if a reduced
number of observations is carried out, then a Student’s t-distribution shall be used, and the calculated
experimental standard deviation of the mean shall be adjusted based on the number of observations made
(see A.2.1.2). For both normal distribution and t-distribution, a coverage factor (divisor) k = 1 shall be used
for computing the contributor’s value (see A.2.1.1 and A.2.1.2).
For a given influence quantity, the decision between using normal or t-distribution for the Type A analysis
depends on the acceptable level of error in the resulting uncertainty budget. The expanded uncertainty shall
be first calculated using the contributor’s value derived by assuming a normal distribution (1 standard
deviation; see A.2.1.1) and then by using the standard deviation adjusted with the correction factor
corresponding to the number of observations performed (see Table A.2 and A.2.1.2). The difference
between the two, expressed as a percentage of the later, represents the error in the expanded uncertainty
budget.
For less than five observations, it is recommended not to use a Type A estimation, but to perform a Type B
analysis instead.
In the context of a Type A evaluation, repeated observations may consist of a set of repeated measurements
or of using a number of previous calibration certificates of a given measurement equipment to calculate the
standard deviation of a specific parameter. Care shall be taken that only the contributor evaluated is varied
during the repeated observations, while all others are kept constant, to avoid double or multiple counting of
contributors in the resulting overall uncertainty budget. As such, a Type A analysis based on previous
calibration certificates cannot usually be used for contributors that are or may be susceptible to aging
and/or drift (except when the aging and/or drift effects are not accounted for through other contributors in
the overall uncertainty budget).
19
ANSI C63.23-2012
Uncertainty
5.6.2 Uncertainty contributors common to all conducted and radiated emission EMC
measurements—Type A evaluation
5.6.2.1 Receiver reading
Receiver readings will vary for different reasons, such as measuring system instability and meter scale
interpolation errors. The estimate of this parameter is the mean of many readings, with a standard
uncertainty given by the experimental standard deviation of the mean (k = 1).
5.6.2.2 Attenuation of the connection between signal transducer and receiver
The attenuation of the cable used to connect the signal transducer (LISN, current probe, or antenna) to the
receiver can be measured using a separate set of instruments such as a network analyzer or a spectrum
analyzer with a signal generator. The standard deviation of the repeated measurements is the contribution
value.
5.6.2.3 Receiver sine-wave voltage accuracy
The Type B evaluation of this contributor is discussed in 5.2.3. An estimate of the magnitude for receiver
sine-wave voltage accuracy can be found in the calibration certificate. Multiple receiver calibrations will
result in multiple results for the sine wave accuracy. These calibration results can be used as the basis for a
Type A evaluation having a normal distribution. Because the sample size is very likely to be relatively
small, a Student’s t-distribution should be used. An alternative approach would be for a test laboratory to
conduct an analysis of variance (ANOVA) of the sine wave accuracy, using the calibration process defined
in the calibration manual of the instrument manufacturer and, if possible, including multiple units of the
same equipment type in the study to avoid biasing.
NOTE—If the test laboratory does not use multiple units for the ANOVA, then a bias may be present in the Type A
evaluation and not be truly representative of the contributor. Caution should be used if the ANOVA is used such that
multiple units of the same device are used.
5.6.2.4 Receiver pulse amplitude response
The Type B evaluation of this contributor is covered in 5.2.4. It is important to note that this factor is only a
contributor when measuring a noncontinuous wave. During the calibration, continuously repeated pulses
are applied and the response measured for each quasi-peak (QP) and average (AVG) detector bandwidth
setting. The value is recorded. Then, the difference is calculated between the measured amplitude value and
the target for each bandwidth of interest. This procedure is repeated for a number of frequency points per
each detector per each bandwidth setting to verify the receiver compliance with the pulse amplitude
response requirements of CISPR 16-1-1:2007. If each one of these measurements is repeated a number of
times, then the results will form a small statistical data set, which can then be used in a Type A analysis to
compute the value of this contributor.
Because the receiver pulse response is subject to aging and temperature drift, this procedure can only be
used if the repeated calibration cycles are all performed within a relatively short period of time (not by
using previous years’ calibration reports).
20
ANSI C63.23-2012
Uncertainty
5.6.2.5 Receiver pulse response variation with repetition frequency
The Type B evaluation of this contributor is covered in 5.2.5. It is important to note that this is only a
contributor to the measurement uncertainty of broadband signals and that it is not a factor in uncertainty of
measurements of narrowband emissions. Similar to the pulse amplitude response, this contributor is
assessed during the normal instrument calibration. The response of the instrument to different pulse
repetition rates over numerous calibration cycles can be used as small sample Type A evaluation. The
calibration report for the instrument will provide the amplitude error at each pulse repetition rate for each
QP and AVG bandwidth setting. These data, taken from multiple calibration cycles, form a small statistical
data set and the standard deviation can be computed at each pulse rate for each bandwidth. The greatest
resulting standard deviation (i.e., of the different pulse rates) should be used for each bandwidth of interest.
Because the receiver pulse response is subject to aging and temperature drift, this procedure can only be
used if the repeated calibration cycles are all performed within a relatively short period of time (not by
using previous years’ calibration reports).
5.6.2.6 Receiver noise floor proximity
This contributor is negligible if the receiver noise floor is more than 10 dB below the signal level that
corresponds to the applicable emissions limit. If this margin is less than 10 dB, then the following
procedure can be employed for estimating the contribution to the conducted or radiated emissions
measurement uncertainty budget due to the receiver noise floor proximity. The example provided is given
for radiated emissions within 30 MHz to 1000 MHz, but a similar procedure can be used for conducted
emissions (below 30 MHz) or for radiated emissions above 1 GHz.
To perform this analysis, a reference EUT is required, which exhibits stable (in both level and frequency)
radiated emissions at or close to the applicable limit, at a set of frequencies evenly spread out across the
30 MHz to 1000 MHz band.
Repeated measurements shall be performed at all frequencies radiated by the reference EUT, with each
measurement (at each frequency) taken twice: with the EUT ON and with the EUT OFF. For each
measurement, the ratio of [signal + noise (EUT ON)] to the [noise (EUT OFF)] shall be computed. The error
in radiated emissions measurement due to the receiver noise floor proximity can be found using the
calculated (S + N)/N value and the plots from Moulthrop and Muha [B19]. This procedure shall be repeated
for a minimum of five iterations, and the standard deviation shall be calculated for each test frequency. The
highest standard deviation thus obtained, from across all investigated frequency points, is the contribution
value.
5.6.2.7 Mismatch effects between signal transducer and receiver
There is no efficient means to assess this contributor with a Type A method because of the following
reasons:
⎯ For conducted emissions, the impedance at the LISN receiver port will be affected by each
unique product connected for test.
⎯ For radiated emissions, the impedance at the antenna port will be affected by the test site.
A Type B evaluation of this contributor should be used instead.
21
ANSI C63.23-2012
Uncertainty
5.6.2.8 Preamplifier gain variation
This contributor needs to be considered if a preamplifier external to the receiver is used during the
measurements. It also needs to be considered when a preamplifier internal to the receiver is used and when
that preamplifier’s gain variation is not taken into account during the receiver calibration routine (or when
the receiver calibration routine is not performed before the measurements).
A Type A evaluation of this contributor can be performed by repeatedly measuring the preamplifier gain,
computing the standard deviation of the obtained results at each measured frequency, and taking the worst-
case (largest) such standard deviation across the applicable frequency range as the contributor’s value.
Alternatively, the Type A analysis can be performed by using the previous year’s calibration certificates. A
minimum of 5 previous years’ (most recent) calibration reports shall be used; otherwise, a Type B analysis
shall be performed. If the Type A analysis is based on old calibration reports and not on repeated
measurements, then the effect of aging on the preamplifier gain will also be accounted for, and in this case,
a separate contributor for the aging effect shall not be used in the uncertainty budget.
5.6.2.9 Instability of preamplifier
This contributor needs to be considered if a preamplifier external to the receiver is used during the
measurements. It also needs to be considered when a preamplifier internal to the receiver is used and when
that preamplifier’s gain variation is not taken into account during the receiver calibration routine (or when
the receiver calibration routine is not performed before the measurements).
Instability of preamplifier gain is due to two contributors: aging and temperature drift.
The first one can only be evaluated through a Type A analysis of the previous years’ calibration reports
(see 5.6.2.8) and not through repeated measurements. In this case, only one contributor will be used in the
uncertainty budget to account for both preamplifier gain variation and instability of preamplifier gain due
to aging.
The temperature drift effect need only be considered if a large variation in ambient temperature exists
between different measurements (usually only for radiated emissions measurement on an OATS). This
contributor cannot be evaluated from previous calibration certificates. A Type A analysis of the
preamplifier gain temperature drift may be performed by taking repeated measurements, at various
temperatures within the expected temperature range and at different frequencies within the applicable
frequency range, calculating the standard deviation of the obtained results at each measured frequency and
taking the worst-case (largest) such standard deviation across the applicable frequency range as the
contributor’s value. Note that a minimum of five measurements must be taken at each frequency, at
different temperatures, and the selected temperature values must be the same for all frequency points and
must be evenly spread across the expected temperature range.
When a statistically representative number of measurements (i.e., minimum five) is taken at each frequency
and temperature combination, then the resulting contributor’s value will account for both preamplifier gain
variation and instability of preamplifier gain due to drift.
22
ANSI C63.23-2012
Uncertainty
5.6.3 Uncertainty contributors for conducted emission measurements at a mains port

using an LISN—Type A evaluation
5.6.3.1 General
In addition to the contributors included in 5.6.2, uncertainty quantities to be considered for conducted
disturbance measurements at a mains port using an LISN are as follows.
5.6.3.2 LISN voltage division factor
This contributor can be assessed by measuring the voltage division factor repeatedly and computing the
standard deviation of those repeated measurements.
5.6.3.3 LISN impedance
This contributor can be assessed by repeating the calibration procedure as specified in ANSI C63.4 and
computing the standard deviation of those repeated measurements.
5.6.4 Uncertainty contributors for radiated emission measurements in the frequency range
30 MHz to 1000 MHz on an OATS or in an SAC—Type A evaluation
5.6.4.1 General
In addition to the contributors included in 5.6.2, uncertainty quantities to be considered for radiated
emission measurements in the frequency range 30 MHz to 1000 MHz on an OATS or in a SAC are as
follows.
5.6.4.2 Antenna factor accuracy
A Type A evaluation of the contribution of the antenna factor uncertainty can be made by computing the
standard deviation of the previously measured antenna factors (minimum five measurements).
This analysis is to be performed for each frequency of calibration. The maximum standard deviation
obtained from across all analyzed frequency points shall be used as the contributor value.
Because the antenna factor is subject to temperature drift, this procedure can be used only if the repeated
calibration cycles are all performed within a relatively short period of time (not by using previous years’
calibration reports). However, if measurements are always performed within a SAC, then a Type A analysis
based on previous years’ calibration reports can be used instead of the repeated measurements because the
temperature inside a SAC is controlled and relatively constant.
5.6.4.3 Antenna factor frequency interpolation
There is no efficient means to assess this contributor with a Type A method. A Type B evaluation of this
contributor should be used instead.
23
ANSI C63.23-2012
Uncertainty
5.6.4.4 Antenna factor variation with height
The antenna calibration processes for Standard Site Method (SSM) and Reference Antenna Method (RAM)
include this variable. Consequently, a series of at least five antenna calibrations of the same type (e.g., SSM
or RAM) will provide a sufficient data set for a Student’s t-distribution.
NOTE 1—Repeated calibrations will account not only for this contributor but also for the antenna factor (AF)
accuracy, directivity, and phase center location contributors. As such, if this Type A analysis is performed, then the
resulting contributor's value shall be used in the overall budget to represent all these contributors. It should be noted
that the other contributors should not be included in the overall budget to ensure the contributor is not counted more
than one time.
NOTE 2—The reference antenna method in ANSI C63.5-2006 [B1] does not require a receive antenna height scan. It is
recommended but not mandatory (see subclause 6.1 of ANSI C63.5-2006). If the antenna calibration does not include a
height scan, then repeating this method in a Type A analysis will not account for this contributor to uncertainty
resulting in the need for a Type B evaluation of this contributor.
If a complete data set is not available, then a Type B evaluation of this contributor should be used instead.
5.6.4.5 Antenna directivity
The antenna calibration processes for Standard Site Method and Reference Antenna Method include this
variable. Consequently, a series of at least five antenna calibrations of the same type (e.g., SSM or RAM)
will provide a sufficient data set for a Student’s t-distribution.
NOTE 1—Repeated calibrations will account not only for this contributor but also for the AF accuracy, directivity, and
phase center location contributors. As such, if this Type A analysis is performed, then the resulting contributor's value
shall be used in the overall budget to represent all these contributors. It should be noted that the other contributors
should not be included in the overall budget to ensure the contributor is not counted more than one time.
NOTE 2—The reference antenna method in ANSI C63.5-2006 [B1] does not require a receive antenna height scan. It is
recommended but not mandatory (see subclause 6.1 of ANSI C63.5-2006 [B1]). If the antenna calibration does not
include a height scan, then repeating this method in a Type A analysis will not account for this contributor to
uncertainty resulting in the need for a Type B evaluation of this contributor.
NOTE 3—The Standard Site Method does not include testing in the vertical polarization. Therefore, if this method is
used for antenna calibration, then either the laboratory needs to request data in the vertical polarization or a Type B
evaluation for the vertical polarization is to be used.
NOTE 4—The Reference Antenna Method does not specify a polarization. Therefore, if the laboratory desires to use a
Type A analysis of this contributor, then the laboratory must request data in both polarizations; otherwise, a Type B
evaluation is to be used.
5.6.4.6 Antenna phase center location
The antenna calibration processes for Standard Site Method and Reference Antenna Method include this
variable. Consequently, a series of at least five antenna calibrations of the same type (e.g., SSM or RAM)
will provide a sufficient data set for a Student’s t-distribution.
NOTE—Repeated calibrations will account not only for this contributor but also for the AF accuracy, directivity, and
phase center location contributors. As such, if this Type A analysis is performed, then the resulting contributor's value
shall be used in the overall budget to represent all these contributors. It should be noted that the other contributors
should not be included in the overall budget to ensure the contributor is not counted more than one time.
24
ANSI C63.23-2012
Uncertainty
5.6.4.7 Antenna cross-polarization response
5.6.4.8 Antenna balance (antenna symmetry)
Perform at least five measurements, which will provide a sufficient data set for a Student’s t-distribution, of
balance of the test antenna using the procedure described in ANSI C63.5. The standard deviation of the
error contribution shall be used as the Type A value.
5.6.4.9 Test site contribution
The contribution to measurement uncertainty from the test site and system can be estimated using a Type A
evaluation. This evaluation requires performing the volumetric normalized site attenuation as described in
ANSI C63.4. This evaluation requires five measurements that are taken at different locations on the ground
plane for each geometry and antenna. The standard deviation of these five values at each frequency
represents the worst-case contribution to uncertainty that can be associated with the site imperfections,
cable lay variations, distance measurement errors, antenna height start/stop errors, antenna travel
smoothness (directivity), and height versus frequency variations. Amplifier gain stability would also be
included in this value if the same amplifier were used during calibrations. The standard deviation in the
data of the five locations shall be calculated for each frequency. The contribution to measurement
uncertainty is then the maximum value of these standard deviations for the geometry of interest (for
example, 2 m horizontal transmitting height and 10 m distance for standard calibrations). A Student’s
t-distribution is assumed with n = 5 and a 95% confidence interval, yielding a coverage factor t = 2.78.
5.6.4.10 Separation between equipment under test and measurement antenna
5.6.4.11 Height of table supporting the equipment under test
5.6.4.12 Effect of table material supporting the equipment under test
CISPR 16-1-4:2010 contains procedures for evaluation of the EUT support material. A Type A contributor
can be determined by conducting repeated measurements of this method. This analysis is to be performed
for each frequency range measured. The maximum standard deviation obtained from across all analyzed
frequency points shall be used as the contributor value.
25
ANSI C63.23-2012
Uncertainty
5.6.5 Uncertainty contributors for radiated emission measurements in the frequency range
above 1000 MHz—Type A evaluation
The same uncertainty terms that are used below 1000 MHz can also be used above 1000 MHz. See 5.6.4
for a listing of these uncertainty terms.
5.7 Example spreadsheets
Table 9 and Table 10 are based on example spreadsheets that can be accessed at the following URL:
http://standards.ieee.org/downloads/C63/. The example spreadsheets are the combined result of Type A and
Type B analyses of contributors. Each contributor is entered on a line in the spreadsheet. Then, based on
the particular parameter, the value is entered in column C, the type of contributor is entered in column B,
and the coverage factor (divisor) is assigned in column G. The remaining information is automatically
populated, and the standard uncertainty and the expanded uncertainty are calculated.
26
ANSI C63.23-2012
American National Standard Guide for Electromagnetic Compatibility—Computations and Treatment of Measurement Uncertainty
Table 9 —Conducted emissions—LISN 150 kHz to 30 MHz

Description of Type Parameter Parameter Distribution Normal Divisor Standard Effective degrees Sensitivity Computed Computed
uncertainty A/B uncertainty units distribution (k) deviation in of freedom (for coefficient uncertainty in uncertainty in
contributor limits level of parameter Type A enter the (normalization analysis units analysis units
confidence units number of factor) squared
measurements
minus 1, for Type
B leave blank)
Receiver reading A 2.60E-01 dB Std Dev 95.00% 1 2.60E-01 4 1 0.26 0.068
Attenuation of
the connection
B 3.80E-01 dB Rectangular 100.00% 1.732 2.19E-01 1 0.22 0.048
between antenna
and receiver
AMN voltage
B 2.50E-01 dB Normal 95.45% 2 1.25E-01 1 0.13 0.016
division factor
Receiver sine-
wave voltage B 2.00E+00 dB Rectangular 100.00% 1.732 1.15E+00 1 1.15 1.333
accuracy
Receiver pulse
amplitude B 1.50E+00 dB Rectangular 100.00% 1.732 8.66E-01 1 0.87 0.750
response
Receiver pulse
response
variation with B 1.50E+00 dB Rectangular 100.00% 1.732 8.66E-01 1 0.87 0.750
repetition
frequency
Noise Floor
B 0.00E+00 dB Normal 95.45% 2 0.00E+00 1 0.00 0.000
proximity
Mismatch
effects between
B 5.80E-01 dB U-Shaped 95.45% 1.414 4.10E-01 1 0.41 0.168
antenna port and
receiver
AMN
B 2.60E+00 dB Triangular 100.00% 2.449 1.06E+00 1 1.06 1.127
impedance
Standard uncertainty 2.06

Expanded uncertainty [based on k factor of 2 (95% confidence level)] 4.13
27
ANSI C63.23-2012
American National Standard Guide for Electromagnetic Compatibility—Computations and Treatment of Measurement Uncertainty
Table 10 —Radiated emissions—biconical antenna 30 MHz to 200 MHz, vertical polarization

Description of Type Parameter Parameter Distribution Normal Divisor Standard Effective degrees of Sensitivity Computed Computed
uncertainty A/B uncertainty units distribution (k) deviation in freedom (for Type A coefficient uncertainty in uncertainty in
contributor limits level of parameter enter the number of (normalization analysis units analysis units
confidence units measurements minus factor) squared
1, for Type B leave
blank)
Receiver reading A 2.11 dB Std Dev 68.27% 1 2.11 4 1 2.11 4.464
Attenuation of the
connection between B 0.10 dB Normal 95.45% 2 0.05 1 0.05 0.003
antenna and receiver
Biconical antenna
B 0.40 dB Normal 95.45% 2 0.20 1 0.20 0.040
factor
Receiver sine-wave
B 2.00 dB Rectangular 100.00% 1.732 1.15 1 1.15 1.333
voltage accuracy
Receiver pulse
B 1.50 dB Rectangular 100.00% 1.732 0.87 1 0.87 0.750
amplitude response
Receiver pulse
response variation
B 1.50 dB Rectangular 100.00% 1.732 0.87 1 0.87 0.750
with repetition
frequency
Noise floor
B 0.50 dB Normal 95.45% 2 0.25 1 0.25 0.063
proximity
Mismatch effects
between antenna port B 0.36 dB U-Shaped 1 Std Dev 1.414 0.25 1 0.25 0.065
and receiver
Antenna factor
frequency B 0.25 dB Rectangular 100.00% 1.732 0.14 1 0.14 0.021
interpolation
Antenna factor
B 0.50 dB Rectangular 100.00% 1.732 0.29 1 0.29 0.083
variation with height
Antenna directivity B 0.50 dB Rectangular 100.00% 1.732 0.29 1 0.29 0.083
Antenna phase center B 0.00 dB 100.00% 0.00 0.000
Antenna cross-
B 0.00 dB 100.00% 0.00 0.000
polarization response
Antenna balance B 0.90 dB Rectangular 100.00% 1.732 0.52 1 0.52 0.270
Test site B 2.84 dB Triangular 100.00% 2.449 1.16 1 1.16 1.345
Separation between
EUT and
B 0.10 dB Rectangular 100.00% 1.732 0.06 1 0.06 0.003
measurement
antenna
Height of table
B 0.10 dB Normal 95.45% 2 0.05 1 0.05 0.003
supporting the EUT
Standard uncertainty 3.05

Expanded uncertainty [based on k factor of 2 (95% confidence level)] 6.09
28
ANSI C63.23-2012
Uncertainty
Annex A
(informative)
Aspects of measurement uncertainty theory
A.1 Development of uncertainty estimates
A.1.1 General
Uncertainty is a parameter associated with the result of a measurement. It characterizes the dispersion of
the values that could be reasonably attributed to the measurand. The uncertainty of a measurement is
always an estimate. It may be arrived at by either computation or estimates based on experience. Because
the computations are statistical in nature, the uncertainty computed remains an estimate: a good estimate,
perhaps, but yet only an estimate.
The following terms are commonly used in estimating measurement uncertainty:
⎯ Standard uncertainty, the symbol of which is us

⎯ Combined standard uncertainty, with symbol uc
⎯ Sensitivity coefficients, with symbol ci
⎯ Coverage factor, with symbol k
⎯ Expanded uncertainty, with symbol U
A.1.2 Combined standard uncertainty
The combined standard uncertainty is calculated using the law of propagation of uncertainty (see ISO/IEC
Guide 98-3:2008 (GUM) [B16]). It represents the estimated standard deviation of a measurement result,
which is obtained from the values of several contributors. It is equal to the positive square root of a sum of
terms that are the weighted variances and covariances of all contributors. Equation (A.1) is an example (for
the case where all contributors are uncorrelated and independent from one another):
uc = c12 us12 + c2 2 us 2 2 + cn 2 usn 2 (A.1)
When selecting components or contributors, care must be taken to avoid double-counting of uncertainty
components. For example, if a component is already included in a Type A evaluation, then it should not be
included again separately as a Type B component [see ANSI/NCSL Z540-2:1997 [B4] or the
ISO/IEC Guide 98-3:2008 (GUM) [B16]]. Any bias that is known and used for correcting the measurement
result should not be included in the uncertainty budget.
A useful term, but one not defined officially in the GUM, is “uncertainty driver.” An uncertainty driver is
an influence quantity that is much larger (approximately two or more times) than most other contributors
affecting the value of the measurand. It effectively sets the minimum possible value of the calculated
combined standard uncertainty of the measurement [i.e., it “drives” (or dominates) the measurement
uncertainty more than all other contributors]. In this case, special consideration must be given to the
applicability of the calculation method described in this document and especially to the correct coverage
factor value k used in computing the expanded uncertainty. This is due to the fact that the central limit
29
ANSI C63.23-2012
Uncertainty
theorem may no longer be applied, and the probability distribution of the “uncertainty driver” will also
“drive” the distribution of the resulted expanded uncertainty. See Annex C of UKAS M 3003:2007 [B22]
for details.
A.1.3 Sensitivity coefficient
In some cases, the input quantity to the process may not be in the same units as the output quantity. For
example, the uncertainty in the measurement distance on an open site should be converted to the
uncertainty in the received signal level. In this case, the input quantity is length, but the output quantity is
electrical field strength. It is therefore necessary to relate the output quantity (y) to the input quantity (xi).
The conversion from one unit to the other is the mathematical expression relating the two quantities. If the
mathematical expression is linear (i.e., an algebraic expression with no terms raised to a power different
than unity), then the sensitivity coefficient is unity. In the earlier example, if the measurement distance is in
the far field of the source, then the field strength is inversely proportional to the distance, so that the
fractional error in measuring the separation distance results in a fractional error in the level of the field
strength. On the other hand, if the measurement distance is within the near field of the source, then the field
strength is inversely proportional to a higher order of the distance (e.g., the square of the distance or the
cube of the distance, or some fractional exponent greater than unity). In this case, the sensitivity coefficient
is equal to the exponent of distance that is applicable.
For a measurement position in the far field from the source:
E = md n (A.2)
where
E is the E-field level (in linear units, not logarithmic units)

m is the proportionality constant
d is the measurement distance
n is the exponent of the change of E with respect to distance (n = −1 in the far field)
The exponent n of the distance is the sensitivity coefficient, usually shown as ci in uncertainty expressions.
Taking 20 times the logarithm of the Equation (A.2) gives:
20 log E = 20 log m + 20n log d (A.3)
A small change Δd causes a small change ΔE, giving:
20 log(E ± ΔE ) = 20 log m + 20n log(d ± Δd ) (A.4)
Subtracting 20 log E from both sides gives:
 E ± ΔE  m  d ± Δd 
20 log  = 20 log  + 20n log 
 E  m
   d  (A.5)
And simplifying:
 ΔE   Δd 
20 log1 ±  = 20n log1 ±  (A.6)
 E   d 
30
ANSI C63.23-2012
Uncertainty
The sensitivity coefficient is ci = n. At distances from the EUT that are in the far field, n = −1.
At 30 MHz or above, the 10 m measurement distance is in the far field so that n = −1; thus:
 ΔE   Δd 
20 log1 ±  = −20 log1 ±  (A.7)
 E   d 
In terms of relative errors in the distance and the field strength:
20 log(1 ± ε E ) = −20 log(1 ± ε d ) (A.8)
Thus, if the error in setting the antenna at 10 m from the EUT is 5 cm, then the relative error in d is:
0.05
εd = = 0.005 (A.9)
10
and the error in the field strength in dB becomes:
 − 0.043 dB
ε E = −20 log(1 ± ε d ) =  (A.10)
+ 0.044 dB
dB
A.1.4 Bias or systematic effect
Bias or systematic effect is common in EMC measurements. It takes the form of correction factors (e.g.,
cable loss), conversion factors (e.g., antenna factor), and bias in specific instrumentation. Each systematic
effect has two elements, a known or approximately known or expected value µ and a random variation
about the expected value uµ. The expected value is used as a direct correction of the measurand and the
random variation is combined into the uncertainty of the measurement in Equation (A.1).
A.1.5 Expanded uncertainty
Expanded uncertainty defines an interval about the result of a measurement expected to encompass a large
fraction of possible values of the result. This interval is based on a coverage factor k. This coverage factor
is usually 2 or 3, but it may have other values. A few values of k related to the confidence level are shown
in Table A.1, assuming a normal (Gaussian) probability distribution for the expanded uncertainty (which,
due to the central limit theorem, is usually true, unless an “uncertainty driver” of other than normal
distribution exists).
31
ANSI C63.23-2012
Uncertainty
Table A.1—Relationship between confidence level and coverage factor

for normal distribution
Confidence level, % 50 68.27 80 90 95.45 99.73 99.99
k 0.675 1 1.28 1.645 2 3 3.89
A.2 Application of distributions in Type A and Type B evaluations
A.2.1 Application of distributions in Type A evaluations
A.2.1.1 Normal distribution
The normal (Gaussian or “bell curve”) distribution is commonly applied to Type A evaluations of additive
data.8
It is assigned to standard uncertainty, combined standard uncertainty, and expanded uncertainty.
The first step in a Type A evaluation consists of calculating the arithmetic mean of the repeated
observations:
n
q = qj (A.11)
j =1
where:
q is the mean value of the repeated observations

qj is the value of the jth observation
n is the number of observations
The experimental standard deviation can then be calculated as:
 (q )
n
( )
s qj =
1
n −1
j −q (A.12)
j =1
The standard uncertainty (the contribution to the measurement uncertainty budget due to an influence
quantity that is evaluated using a Type A statistical method) is expressed as 1 standard deviation (or error)
of the mean value:
us = s q =( ) s(qn ) j
(A.13)
The divisor used in the uncertainty budget calculation is in this case k = 1, because the contributor’s value
obtained from Equation (A.13) is already in terms of 1 standard deviation.
8
It is incorrect to use percentages with the normal distribution. Percentages are not additive unless converted to their base units, and
the correct mean value of a group of percentages is the harmonic mean, rather than the arithmetic mean of the normal distribution. The
errors or variations used with the normal distribution must be additive (e.g., decibels or raw linear variations such as ±50 µV/m)
(around a specific value such as 600 µV/m).
32
ANSI C63.23-2012
Uncertainty
NOTE 1—The previous formulas can only be applied when the number of observations is sufficiently large such that
they may be considered to be a statistically representative sample set of the entire population of possible observations
and when the observations are independent from each other (no correlation exists between different observations).
NOTE 2—The quantity ν = n – 1 from the denominator of Equation (A.12) is called number of degrees of freedom, and
it represents the number of independent observations minus the number of parameters derived from these observations.
When the observations are not obtained from repeated measurements (for example, when old calibration data from a
number of previous years is used for the Type A analysis), then ν = n shall be used in Equation (A.12). See
IEC 61000-1-6:2012 [B13].
A.2.1.2 Student’s t-distribution
Student’s probability distribution, or the t-distribution, arises in the problem of estimating the mean of a
normally distributed population when the sample size is small. In practice, the standard deviation of the
population is always unknown and must be estimated from the data.
Suppose there is a random sample set of size n drawn from a normal-distributed population with mean μ
and standard deviation σ. Let x denote the sample mean and s the sample standard deviation. Then, the
quantity:
x−μ
t= (A.14)
s/ n
has a t distribution with n – 1 degrees of freedom. The variance (the standard deviation squared) is then:
( )
1 n
s2 = xi − x (A.15)
n − 1 i =1
Note that there is a different t distribution for each sample set size; in other words, it is a class of
distributions. When specifying a t distribution, the degrees of freedom must also be specified. The degrees
of freedom for this t statistics come from the sample standard deviation s in the denominator of
Equation (A.14).
The t density curves are symmetric and bell shaped like the normal distribution and have their peak at zero.
However, the spread is more than that of the standard normal distribution. This is due to the fact that in
Equation (A.14), the denominator is s rather than σ. Because s is a random quantity varying with various
samples, the variability in t is larger, resulting in a larger spread.
The larger the degrees of freedom, the closer the t-distribution is to the normal distribution. This reflects the
fact that the standard deviation s approaches σ for large sample set size n.
The contributor’s value of a t-distributed influence quantity (normal distributed quantity for which a
sufficiently large sample size cannot be obtained) can be calculated by using Equation (A.11) and
Equation (A.12), and then by adjusting the obtained experimental standard deviation by an appropriate
coefficient, h(n) , which accounts for the limited sample size:
h(ν) s (q j )
us = h(ν) s (q ) = (A.16)
n
The coefficient h(ν) depends on the number of degrees of freedom, ν = n – 1 (see IEC 61000-1-6:2012
[B13]):
33
ANSI C63.23-2012
Uncertainty
ìï
ïï
ïï6.48 ν =1
ï
h(ν) = ïí2.20 ν=2 (A.17)
ïï
ïï ν
ïï ν³3
ïî ν - 2
The values of η (ν ) are calculated as the ratio between the upper critical values of the Student’s
t-distribution for ν degrees of freedom and an infinite degrees of freedom, respectively, for a 95 %
confidence level. As the number of observations increases, the number of degrees of freedom approaches
infinity and η (ν ) approaches unity. A selection of values for η (ν ) is given in Table A.2, for various
degrees of freedom.
The divisor used in the uncertainty budget calculation is in this case k = 1 because the contributor’s value
obtained from Equation (A.16) is already in terms of 1 standard deviation.
Table A.2—Values of the expansion coefficient η(ν)

ν 1 2 3 4 5 6 7 8 9
η(ν) 6.48 2.20 1.73 1.41 1.29 1.22 1.18 1.15 1.13
ν 10 11 12 13 14 19 29 49 99
η(ν) 1.12 1.11 1.10 1.09 1.08 1.06 1.04 1.02 1.01
A.2.2 Applications of distributions in Type B evaluations
A.2.2.1 General
The distributions described in A.2.2.2 through A.2.2.7 are a method to put Type A and Type B evaluations
of uncertainty on the same “footing.” Their equivalent descriptive statistics may be estimated from
calculations using statistically sound approximations. This allows combining, with appropriate weighting,
uncertainties from various sources via the law of propagation of uncertainty. The resulting combined
standard uncertainty is considered normally distributed; thus, the expanded uncertainty can maintain its
prediction of confidence level.
A.2.2.2 Normal distribution
The most common example of a normal-distributed influence quantity in a Type B analysis consists of
using data from a calibration certificate. For instance, the antenna factor, obtained from the calibration
certificate, is used in computing the radiated emission field strength level by adding its value (in dB/m) to
the cable loss (in dB) and to the signal amplitude measured by the receiver (in dBµV). The uncertainty of
calibration, also obtained from the calibration certificate, is then a contributor to the radiated emissions
measurement uncertainty budget.
The value of a normal-distributed uncertainty (such as that obtained from a calibration certificate) may be
reported either in terms of 1 standard deviation or as an expanded uncertainty. In the latter case, the
reported value must be accompanied by a statement of the confidence level and the value of the coverage
factor k used in computing the expanded uncertainty. The contributor’s value is then the reported
uncertainty value, and the divisor is k (for values reported as 1 standard deviation, the divisor is unity). See
Table A.1 for a set of more frequently used values of k.
34
ANSI C63.23-2012
Uncertainty
A.2.2.3 Student’s t-distribution
As stated in A.2.1.2, when estimating the standard deviation of a normal-distributed quantity from a
reduced set of observations, a t-distribution must be used. Type B evaluations of normal-distributed
quantities are most likely based on data from calibration certificates, the vast majority of which report an
expanded uncertainty calculated based on a normal distribution. However, if the expanded uncertainty is
based on a t-distribution rather than a normal distribution, it has to be accompanied by additional
information, like the number of degrees of freedom (ν), confidence level (p), and the coverage factor (kp)
used in the calculation of the reported expanded uncertainty value. If any of these three parameters is not
reported, then it can be derived from the other two by using the t-distribution of Table A.3 (reproduced
from UKAS M 3003:2007 [B22] and ISO/IEC Guide 98-3:2008 [B16]).
If the reported number of degrees of freedom is not listed in Table A.3, then it will be necessary to
interpolate between the kp values given in the table. For ν > 3, linear interpolation is sufficient; for lower
values of ν, higher order interpolation should be used. Alternatively, the next lower value of ν listed in
Table A.3 may be used, which would give a worst-case estimation of the uncertainty value.
The contributor’s value to be used in the overall uncertainty budget is in this case the reported expanded
uncertainty, and the divisor is kp, as determined from Table A.3.
Table A.3—t-Distribution table (kp as a function of ν and p)

Degrees of kp as a function of the number of effective degrees of freedom and confidence level
freedom, ν p = 68.27% p = 90% p = 95% p = 95.45% p = 99% p = 99.73%
1 1.84 6.31 12.71 13.97 63.66 235.80
2 1.32 2.92 4.30 4.53 9.92 19.21
3 1.20 2.35 3.18 3.31 5.84 9.22
4 1.14 2.13 2.78 2.87 4.60 6.62
5 1.11 2.02 2.57 2.65 4.03 5.51
6 1.09 1.94 2.45 2.52 3.71 4.90
7 1.08 1.89 2.36 2.43 3.50 4.53
8 1.07 1.86 2.31 2.37 3.36 4.28
9 1.06 1.83 2.26 2.32 3.25 4.09
10 1.05 1.81 2.23 2.28 3.17 3.96
11 1.05 1.80 2.20 2.25 3.11 3.85
12 1.04 1.78 2.18 2.23 3.05 3.76
13 1.04 1.77 2.16 2.21 3.01 3.69
14 1.04 1.76 2.14 2.20 2.98 3.64
15 1.03 1.75 2.13 2.18 2.95 3.59
16 1.03 1.75 2.12 2.17 2.92 3.54
17 1.03 1.74 2.11 2.16 2.90 3.51
18 1.03 1.73 2.10 2.15 2.88 3.48
19 1.03 1.73 2.09 2.14 2.86 3.45
20 1.03 1.72 2.09 2.13 2.85 3.42
25 1.02 1.71 2.06 2.11 2.79 3.33
30 1.01 1.70 2.04 2.09 2.75 3.27
35 1.01 1.70 2.03 2.07 2.72 3.23
40 1.01 1.68 2.02 2.06 2.70 3.20
45 1.01 1.68 2.01 2.06 2.69 3.18
50 1.01 1.68 2.01 2.05 2.68 3.16
100 1.005 1.660 1.984 2.025 2.626 3.077
∞ 1.000 1.645 1.960 2.000 2.576 3.000
A.2.2.4 Rectangular distribution
The rectangular distribution is used where all values in a continuum are equally likely (e.g., an instrument
manufacturer’s specified tolerance).
35
ANSI C63.23-2012
Uncertainty
The true value Xi may, with equal probability, lie anywhere in the interval a+ to a–. From a mathematical
analysis, u2(xi) = (a+ – a–)2 / 12. For symmetrical data, a+ – a– = 2a, and u2(xi) = a2 / 3, or u(xi) = a / √3
(i.e., weighting factor is 1 / √3). If the data are unsymmetrical, then understand why and/or perform a
further analysis as follows. For example, if the manufacturer specifies a tolerance of +2/−3 dB, then find
that the midpoint is −0.5 dB and that a+ – a- = 5 dB. Then the uncertainty is 2.5 / √3 dB with a bias of
−0.5 dB. This bias is additive with any other biases and is transferred to the final result. If the manufacturer
arrived at this unsymmetrical tolerance due to conversion of percentages to dB, then it is usually best
simply to use the largest absolute value as a in the equation for u(xi) and ignore the bias.
This distribution is applied to tolerances or errors stated as a range of possible values without any
indication of the statistics. For example, when an instruction manual gives the amplitude tolerance of an
instrument as ±2 dB, then the rectangular distribution must be applied because the actual measurand could
lie anywhere within the continuum from +2 dB to –2 dB with equal probability. The standard uncertainty
for a measurand with a tolerance of ±2 dB would be us = 2 / √3 ≈ 1.15 dB. This number would be entered
into Equation (A.1), the root sum of squares equation for the propagation of uncertainty, along with other
standard uncertainties of the measurand.
The following analysis shows how the statistics for the rectangular distribution arise. Similar, although
more complicated, analyses were performed for the remaining non-normal distributions.
The midpoint is Xi and the limits are Xi ± a. The area of the distribution must be unity (100% probability);
therefore, the height of the rectangle must be 1 / (2a), and the probability density function is P = 1 / (2a).
The variance is thus:
a
u 2 (xi ) =  x 2 Pdx (A.18)
−a
Substituting P, then:
a
x2
a x3  a3   − a3  a2
u 2 (xi ) =  dx = =   −   = (A.19)
− a 2a 6a −a  6a   6a  3
Therefore:
a
u (xi ) = (A.20)
3
A further example of the rectangular distribution is the result of repeated rolling or throwing of one fair die
as shown in Figure A.1.
0.25
Probability
0.2
0.15
0.1
0.05
0
0 1 2 3 4 5 6 7
Number
Figure A.1—Rectangular probability distribution of one die (µ = 3.5, a = 2.5, us = 1.44)
36
ANSI C63.23-2012
Uncertainty
A.2.2.5 U-shaped distribution
The U-shaped distribution is used where values tend to be clustered near the ends of a continuum (e.g., the
uncertainty associated with VSWR). The true value Xi is concentrated near the ends of the interval a+ to a–.
From a mathematical analysis, u2(xi) = (a+ – a–)2 / 8. For symmetrical data, a+ – a– = 2a, and u2(xi) = a2 / 2;
thus, u(xi) = a / √2 (i.e., weighting factor is 1 / √2).
Examples include 5% carbon-slug resistor values and the effects of VSWR on measurement uncertainty.
VSWR effects are symmetrical in linear terms but not symmetrical in decibels. In linear terms, mismatch
uncertainty is given by uM = ±2|ГG||ГL| / √2. Mismatch uncertainty, in decibels, is given by uM = [20
log(1+|ГG||ГL|) − 20 log(1−|ГG||ГL|)]/(2√2), dB. Note that this is a special case of Equation (2), where the
interconnecting network between the source and the load is assumed to be well matched (i.e., |S11|<<1 and
|S22|<<1) and of negligible attenuation (|S21|≈1).
If the VSWR for the connection between two devices in the measurement chain were 2.5:1 for the source
device and 1.8:1 for the load device, then the magnitudes of the reflection coefficients |ГG| (of the source)
and |ГL| (of the load) would be 0.428 and 0.286, respectively. The mismatch uncertainty would be
uM = [20 log(1 + 0.428 × 0.286) − 20 log(1 − 0.428 × 0.286)] / 2√2 = [1.003 – (−1.135)] / 2√2 = 0.756 dB.
One can carry both values into the propagation of uncertainty, but it is customary to use the largest of the
absolute values; in this case, 1.135 dB, which will result in uM = 1.135 / √2 = 0.802 dB.
The U-shaped distribution is shown in Figure A.2. The vertical dashed lines indicate the 1/√2 weighting
(±1 standard deviation around the expected value).
0.5
0.4
Probability
0.3
0.2
0.1
0
-1 -0.5 0 0.5 1
Number
Figure A.2—The U-shaped probability distribution
A.2.2.6 A priori (triangular) distribution
The a priori or triangular distribution is used where values tend to be clustered near the center of a
continuum, as in the uncertainty associated with site attenuation measurements. The true value Xi is known
or reasonably believed to be concentrated near and around the center of the interval a+ to a–. From
mathematical analysis, u2(xi) = (a+ – a–)2/24. For symmetrical data, u(xi) = a/√6 (weighting factor is 1/√6).
Some examples are site attenuation data on a well-behaved open-area test site, and other processes that
produce tightly clustered, centrally located data.
The criterion in ANSI C63.6-1996 [B2] and ANSI C63.4 for the acceptance of an OATS is ±4 dB. Thus, in
finding the standard uncertainty for the OATS as an influence quantity of a radiated emission measurement,
us = 4 / √6 = 1.63 dB. This value is entered into the equation for the propagation of uncertainty as the
influence quantity corresponding to the OATS site attenuation compliance.
37
ANSI C63.23-2012
Uncertainty
The triangular distribution is shown in Figure A.3.
Probability 1/a
of Xi
a– µx a+
µx – a / 2.45 µx + a / 2.45
Figure A.3—Triangular (a priori) distribution
A.2.2.7 Combinations of distributions
Figure A.1 showed the distribution created by rolling or throwing one fair die. Figure A.4 shows the
distribution resulting from rolling two fair dice, and Figure A.5 shows the distribution of the results of
rolling three fair dice.
This process may be repeated for more and more dice, and as can be seen from the examples in Figure A.1,
Figure A.4, and Figure A.5, the more rectangular distributions one combines, the more nearly a normal
distribution results. This may help the reader to understand how the combined standard uncertainty
becomes a normal distribution even though it is made up of non-normal distributions (the central limit
theorem).
0.18
0.16
0.14
Probability
0.12
0.1
0.08
0.06
0.04
0.02
0
1 2 3 4 5 6 7 8 9 10 11 12 13
Number
Figure A.4—Probability distribution resulting from rolling two fair dice
38
ANSI C63.23-2012
Uncertainty
25
20
15
Density
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Num ber
Figure A.5—Probability distribution resulting from rolling three fair dice
A.3 Relationship of log-normal and normal distributions
The normal (Gaussian) distribution was designed and is intended for the analysis of additive data. The log-
normal distribution was designed and is intended for the analysis of multiplicative data. The transformation
from the log-normal distribution to the normal distribution is f(Y) = ln g(X), where f(Y) is the normal
distribution and g(X) is the log-normal distribution. The transform of the log-normal distribution into the
normal distribution is shown in Figure A.6.
0.7
Y = Ln X (Norm.)
f(X) (Lognorm.) 0.6
f(X) Mode, Median, Mean
0.5
f(X) Median 0.4
f f(X) Mean
0.3
f(X) Mode
0.2
0.1
0
–5 –4 –3 –2 –1 0 1 2 3 4 5
X
Y
Figure A.6—Transform of the log-normal distribution to the normal distribution showing

locations of the principal log-normal statistics transformed to the normal distribution
39
ANSI C63.23-2012
Uncertainty
EMC test data are multiplicative in nature. When one makes a measurement, for example, radiated field
strength, the instrument reading (units in µV) is multiplied by the antenna factor (units in 1/m), and then
multiplied by the cable loss (units in V/V). Usually, the measurements are made in terms of decibels, so the
previous example becomes one of addition of decibel values. To the instrument reading (units in dB[µV])
are added the antenna factor (units in dB[1/m]) and cable loss (units in dB). Because the transformation
from the log-normal distribution to the normal distribution is f(Y) = ln g(X), the direct transformation of
levels in linear units is via the same transformation. For example, the instrument reading X in µV is Y = ln
X nepers, and the conversion of nepers to decibels referenced to voltage is 8.686; that is, Y = 8.686 ln X =
20 log X, in dB(µV).
This transformation cannot be correctly used with percentages because they are neither multiplicative nor
additive!
40
ANSI C63.23-2012
Uncertainty
Annex B
(informative)
Bibliography
Bibliographical references are resources that provide additional or helpful material but do not need to be
understood or used to implement this standard. Reference to these resources is made for informational use
only.
[B1] ANSI C63.5-2006, American National Standard for Electromagnetic Compatibility—Radiated

Emission Measurements in Electromagnetic Interference (EMI) Control—Calibration of Antennas (9 kHz
to 40 GHz).
[B2] ANSI C63.6-1996, American National Standard Guide for the Computation of Errors in Open-Area
Test Site Measurements.9
[B3] ANSI/NCSL Z540-1:1994, American National Standard for Calibration—Calibration Laboratories
and Measuring and Test Equipment—General Requirements.
[B4] ANSI/NCSL Z540-2:1997, American National Standard for Expressing Uncertainty — U.S. Guide
to the Expression of Uncertainty in Measurement.
[B5] Carpenter, D., “A demystification of the U-shaped probability distribution,” Proceedings of the
IEEE International Symposium on Electromagnetic Compatibility, Boston, MA, 2003.
[B6] Carpenter, D., “A further demystification of the U-shaped probability distribution,” Proceedings of
the IEEE International Symposium on Electromagnetic Compatibility, Chicago, IL, 2005.
[B7] CISPR/A/838/INF, January 2009, containing CISPR/A/WG1(Dunker-Riedelsheimer- Trautnitz)06-
01, Measurement of FAR similar to CISPR 16-1-4 and site VSWR in the Kolberg FAR of the Federal
Network Agency for Electricity, Gas, Telecommunications, Post and Railway, September 2006
(background material on an estimation of the uncertainty due to results of SVSWR measurements, to be
published before the FDIS).
[B8] CISPR/TR 16-4-1:2009, Specification for Radio Disturbance and Immunity Measuring Apparatus
and Methods—Part 4-1: Uncertainties, Statistics and Limit Modelling—Uncertainties in Standardized EMC
Tests.10
[B9] CISPR/TR 16-4-3:2007, Specification for Radio Disturbance and Immunity Measuring Apparatus
and Methods—Part 4-3: Uncertainties, Statistics and Limit Modelling—Statistical Considerations in the
Determination of EMC Compliance of Mass-Produced Products.
[B10] CISPR 11-2010, Industrial, Scientific, and Medical (ISM) Radio-Frequency Equipment—
Electromagnetic Disturbance Characteristics—Limits and Methods of Measurement.
[B11] CISPR 16-2-3:2006-2007, Specification for Radio Disturbance and Immunity Measuring Apparatus
and Methods—Part 2-3: Methods of Measurement of Disturbances and Immunity—Radiated Disturbance
Measurements.
[B12] CISPR 22-2010, Information Technology Equipment—Radio Disturbance Characteristics—Limits
and Methods of Measurement.
9
ANSI publications are available from the American National Standards Institute (http://www.ansi.org/).
10
CISPR documents are available from the International Electrotechnical Commission (http://www.iec.ch/). They are also available in
the United States from the American National Standards Institute (http://www.ansi.org/).
41
ANSI C63.23-2012
Uncertainty
[B13] IEC 61000-1-6:2012, Electromagnetic Compatibility (EMC)— Part 1-6: General—Guide to the
Assessment of Measurement Uncertainty, Ed. 1.11
[B14] ISO 10012-2003, Measurement Management Systems—Requirements for Measurement Processes
and Measuring Equipment.12
[B15] ISO/IEC 17000-2004, Conformity Assessment—Vocabulary and General Principles.13
[B16] ISO/IEC Guide 98-3:2008, Uncertainty of Measurement—Part 3: Guide to the Expression of
Uncertainty in Measurement (GUM:1995).
[B17] ISO/IEC Guide 99-2007, International Vocabulary of Metrology—Basic and General Concepts and
Associated Terms (VIM).
[B18] Medler, J., “Uncertainty contribution of the EMI test receiver in RF disturbance measurements,”
Proceedings IEEE Asia-Pacific International Symposium on EMC, Beijing, China, 2010.
[B19] Moulthrop, A. A. and Michael S. Muha,, “Accurate measurement of signals close to the noise floor
on a spectrum analyzer,” IEEE Transactions on Microwave Theory and Techniques, pp. 1182–1885,
Nov. 1991.
[B20] Stecher, M., “A detailed analysis of EMI test receiver measurement uncertainty,” Proceedings IEEE
International Symposium on EMC, Montreal, Canada, 2001.
[B21] UKAS LAB 34:2002, The Expression of Uncertainty in EMC Testing.14
[B22] UKAS M 3003:2007, The Expression of Uncertainty and Confidence in Measurement.
[B23] Warner, F. L., “New expression for mismatch uncertainty when measuring microwave attenuation,”
IEEE Proceedings, Part H—Microwaves, Optics and Antennas, vol. 127, Part H, no. 2, Apr. 1980.
11
IEC publications are available from the International Electrotechnical Commission (http://www.iec.ch/). IEC publications are also
available in the United States from the American National Standards Institute (http://www.ansi.org/).
12
ISO publications are available from the ISO Central Secretariat (http://www.iso.org/). ISO publications are also available in the
13
ISO/IEC publications are available from the ISO Central Secretariat (http://www.iso.org/). ISO publications are also available in the
14
UKAS publications are available from the United Kingdom Accreditation Service (http://www.ukas.com).
42

American National Standard Guide For Electromagnetic CompatibilityComputations and Treatment of Measurement Uncertainty

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

American National Standard Guide For Electromagnetic CompatibilityComputations and Treatment of Measurement Uncertainty

Transféré par

Droits d'auteur :

Formats disponibles

American National Standard

Guide for Electromagnetic

Accredited Standards Committee C63® – Electromagnetic Compatibility

Institute of Electrical and Electronics Engineers, Inc.

Approved 3 December 2012

American National Standards Institute

The Institute of Electrical and Electronics Engineers, Inc.

Copyright © 2013 by The Institute of Electrical and Electronics Engineers, Inc.

PDF: ISBN 978-0-7381-8222-3 STD98135

Laws and regulations

Updating of IEEE documents

Daniel Hoolihan, Chair

Dennis Camell, Chair

Mark Arthurs H. R. (Bob) Hofmann Nate Potts

Bob DeLisi, Chair

Tim Harrington Victor Kuczynski Nate Potts

This guide may apply to other C63® standards as appropriate.

4. Basic concepts ............................................................................................................................................ 5

5. EMC measurement uncertainty calculations .............................................................................................. 8

Annex A (informative) Aspects of measurement uncertainty theory ........................................................... 29

Annex B (informative) Bibliography............................................................................................................ 41

ISO 21749:2005, Measurement and Uncertainty for Metrological Applications—Repeated Measurements

measurand: Specific quantity subject to measurement.

4.1 Introduction to measurement uncertainty

4.2 Concepts of uncertainty

4.2.2 Type A evaluation

4.2.3 Type B evaluation

4.2.4 Type A or Type B evaluation

4.2.5 Measurement uncertainty budget calculation

Table 1 —Calculation of contributors to the measurement uncertainty budget

4.3 Application of uncertainty values in conformity assessment

5. EMC measurement uncertainty calculations

Each method is divided into two evaluations, as follows:

Table 2 —List of contributing factors

5.2.1 Receiver reading

5.2.2 Attenuation of the connection between signal transducer and receiver

5.2.3 Receiver sine-wave voltage accuracy

5.2.4 Receiver pulse amplitude response

5.2.5 Receiver pulse response variation with repetition frequency

5.2.6 Receiver noise floor proximity

Table 3 —Receiver input VSWR specifications per CISPR 16-1-1:2010

5.2.8 Preamplifier gain variation

5.2.9 Instability of preamplifier

5.3 Uncertainty contributors for conducted emission measurements at a mains

5.3.2 LISN voltage division factor

5.3.3 LISN impedance

5.4 Uncertainty contributors for radiated emission measurements in the frequency

5.4.2 Antenna factor accuracy

5.4.3 Antenna factor frequency interpolation

5.4.4 Antenna factor variation with height

5.4.5 Antenna directivity

Table 4 —Horizontally polarized radiated emissions from 30 MHz to 200 MHz

Table 6 —Horizontally polarized radiated disturbances from 200 MHz to 1 GHz

Table 7 —Vertically polarized radiated disturbances from 200 MHz to 1 GHz

5.4.6 Antenna phase center location

5.4.7 Antenna cross-polarization response

Cross polarization error is defined as:

where x is the linear response and y is the cross-polarized response in dBµV.

5.4.8 Antenna balance

Table 8 —Estimate of antenna balance contributor

5.4.9 Test site contribution

5.4.10 Separation between equipment under test and measurement antenna

5.4.11 Height of table supporting the equipment under test