NATIONAL RADIO SYSTEMS COMMITTEE NRSC-G201-B NRSC

NATIONAL RADIO SYSTEMS COMMITTEE NRSC-G201-B NRSC
NRSC
GUIDELINE
NATIONAL
RADIO
SYSTEMS
COMMITTEE
NRSC-G201-B
NRSC-5 RF Mask Compliance:
Measurement Methods and Practice
April 2016
NAB: 1771 N Street, N.W.
Washington, DC 20036
Tel: 202-429-5356 Fax: 202-517-1617
1919 South Eads Street
Arlington, VA 22202
Tel: 703-907-4366 Fax: 703-907-4158
Co-sponsored by the Consumer Technology Association and the National Association of Broadcasters
http://www.nrscstandards.org
NRSC-G201-B
NOTICE
NRSC Standards, Guidelines, Reports and other technical publications are designed to serve the public
interest through eliminating misunderstandings between manufacturers and purchasers, facilitating
interchangeability and improvement of products, and assisting the purchaser in selecting and obtaining
with minimum delay the proper product for his particular need. Existence of such Standards, Guidelines,
Reports and other technical publications shall not in any respect preclude any member or nonmember of
the Consumer Technology Association (CTA) or the National Association of Broadcasters (NAB) from
manufacturing or selling products not conforming to such Standards, Guidelines, Reports and other
technical publications, nor shall the existence of such Standards, Guidelines, Reports and other technical
publications preclude their voluntary use by those other than CTA or NAB members, whether to be used
either domestically or internationally.
Standards, Guidelines, Reports and other technical publications are adopted by the NRSC in accordance
with the NRSC patent policy. By such action, CTA and NAB do not assume any liability to any patent
owner, nor do they assume any obligation whatever to parties adopting the Standard, Guideline, Report
or other technical publication.
This Guideline does not purport to address all safety problems associated with its use or all applicable
regulatory requirements. It is the responsibility of the user of this Guideline to establish appropriate safety
and health practices and to determine the applicability of regulatory limitations before its use.
Published by
CONSUMER TECHNOLOGY ASSOCIATION
Technology & Standards Department
1919 S. Eads St.
Arlington, VA 22202
NATIONAL ASSOCIATION OF BROADCASTERS
Science and Technology Department
1771 N Street, NW
Washington, DC 20036
©2016 CTA & NAB. All rights reserved.
This document is available free of charge via the NRSC website at
www.nrscstandards.org. Republication or further distribution of this
document, in whole or in part, requires prior permission of CTA or NAB.
Page 1
NRSC-G201-B
FOREWORD
The NRSC-5 in-band/on-channel (IBOC) Digital Radio Broadcasting Standard specifies IBOC systems for
both the AM and FM bands, including detailed specification of the parameters which must be met by the
radio frequency (RF) signal which is ultimately broadcast by an IBOC facility. The most common
graphical expression of some of these parameters is the so-called “RF mask” which can be thought of as
a template within which the IBOC RF signal must fit.
Given the variety of transmission facility configurations suitable for IBOC signal generation as well as the
need to characterize IBOC RF signals by both equipment manufacturers (during construction and testing
of IBOC transmission equipment) and broadcasters, various methods must be utilized to determine
whether an RF signal is compliant with the RF masks specified by NRSC-5. The purpose of the NRSC
Guideline is to provide background information as well as detailed instructions on the best methods and
practices for determining RF mask compliance for the situations most likely to be encountered by
equipment manufacturers and broadcasters.
The information contained in this NRSC Guideline is the work of the IBOC Standards Development
Working Group (ISDWG), a subgroup of the Digital Radio Broadcasting (DRB) Subcommittee of the
NRSC. At the time of first adoption of this Guideline, the ISDWG was chaired by Dom Bordonaro, Cox
Radio, and the DRB Subcommittee was co-chaired by Mike Bergman, Kenwood Americas Corporation,
and Andy Laird, Journal Broadcast Group. The NRSC chairman at the time of adoption of NRSC-G201
was Milford Smith, Greater Media, Inc.
The NRSC is jointly sponsored by the Consumer Technology Association and the National Association of
Broadcasters. It serves as an industry-wide standards-setting body for technical aspects of terrestrial
over-the-air radio broadcasting systems in the United States.
Page 2
NRSC-G201-B
CONTENTS
1
SCOPE ............................................................................................................................................. 7
2
REFERENCES................................................................................................................................. 7
2.1
NORMATIVE REFERENCES .............................................................................................................. 7
2.2
INFORMATIVE REFERENCES ............................................................................................................ 7
2.3
SYMBOLS AND ABBREVIATIONS ....................................................................................................... 8
2.4
DEFINITIONS .................................................................................................................................. 9
3
BACKGROUND ............................................................................................................................. 11
4
HYBRID FM IBOC MASK COMPLIANCE MEASUREMENTS................................................................... 12
4.1
OVERVIEW ................................................................................................................................... 12
4.2
MODULATING SIGNALS AND COMPLIANCE MEASUREMENTS.............................................................. 13
4.3
OPERATIONAL FACILITIES.............................................................................................................. 13
4.3.1 Common-line facilities ........................................................................................................... 15
4.3.1.1
4.3.1.2
4.3.1.3
4.3.2
Measurement location and sampling method ............................................................................. 15
Instrument configuration ............................................................................................................. 16
Spurious emissions ..................................................................................................................... 16
Separate-line facilities ........................................................................................................... 18
4.3.2.1
4.3.2.2
4.3.2.3
Measurement location and sampling method ............................................................................. 18
Instrument configuration ............................................................................................................. 19
Spurious emissions ..................................................................................................................... 19
4.4
TROUBLESHOOTING (VARIOUS OUT-OF-SPEC CONDITIONS AND POSSIBLE REMEDIES) ....................... 20
4.4.1 Pre-correction and clipping discussion .................................................................................. 21
4.4.2 Evaluating ingress of unwanted signals ................................................................................ 21
4.4.3 Digital Signal Quality Testing ................................................................................................ 23
4.4.3.1
4.4.3.2
4.4.3.3
Modulation Error Ratio (MER)..................................................................................................... 23
Bit Error Ratio (BER) .................................................................................................................. 24
MER vs. BER .............................................................................................................................. 24
5
HYBRID AM IBOC M ASK COMPLIANCE MEASUREMENTS ................................................................. 25
5.1
OVERVIEW ................................................................................................................................... 25
5.2
OPERATIONAL FACILITIES.............................................................................................................. 26
5.2.1 Directional and non-directional antenna systems ................................................................. 27
5.2.2 Measurement locations and sampling methods .................................................................... 27
5.2.2.1
5.2.2.2
5.2.2.3
5.2.2.4
5.2.3
5.2.4
5.2.5
Field measurement ..................................................................................................................... 27
True power measurement at transmitter ..................................................................................... 28
Impact of pattern bandwidth........................................................................................................ 29
Directional coupler measurements not for compliance assessment ........................................... 29
Analog signal and digital sidebands ...................................................................................... 29
Spurious emissions ............................................................................................................... 30
AM IBOC mask compliance measurement procedures ........................................................ 31
5.2.5.1
5.2.5.2
5.2.5.3
Additional legacy mask on full hybrid signal measurement......................................................... 32
Harmonics................................................................................................................................... 32
Testing for instrument-induced harmonic and spurious content.................................................. 32
5.3
TROUBLESHOOTING (VARIOUS OUT-OF-SPEC CONDITIONS AND POSSIBLE REMEDIES) ....................... 32
5.3.1 Digital Signal Quality Testing ................................................................................................ 33
5.3.1.1
5.3.1.2
5.3.1.3
Modulation Error Ratio (MER)..................................................................................................... 33
Bit Error Ratio ............................................................................................................................. 33
MER vs. BER .............................................................................................................................. 34
6
MEASUREMENT METHODOLOGY........................................................................................................ 35
6.1
SPECTRUM ANALYZERS ................................................................................................................ 35
6.1.1 Input levels ............................................................................................................................ 35
6.1.2 Dynamic range ...................................................................................................................... 35
6.1.3 Resolution bandwidth (RBW) and noise bandwidth .............................................................. 37
6.1.4 Detectors ............................................................................................................................... 37
6.1.5 Limit lines ............................................................................................................................... 38
6.2
FM SETUP – SPECTRUM ANALYZERS ............................................................................................. 38
6.2.1 Setting frequency span .......................................................................................................... 38
Page 3
NRSC-G201-B
6.2.2 Resolution bandwidth (RBW) ................................................................................................ 39
6.2.3 Video bandwidth .................................................................................................................... 39
6.2.4 Detector ................................................................................................................................. 39
6.2.5 Reference level...................................................................................................................... 39
6.2.6 Sweep rate and number of sweeps ....................................................................................... 40
6.2.7 Interpreting results ................................................................................................................. 41
6.3
AM SETUP – SPECTRUM ANALYZERS ............................................................................................. 42
6.3.1 Setting frequency span .......................................................................................................... 42
6.3.2 Resolution bandwidth (RBW) ................................................................................................ 42
6.3.3 Video bandwidth .................................................................................................................... 43
6.3.4 Detector ................................................................................................................................. 43
6.3.5 Reference level...................................................................................................................... 43
6.3.6 Sweep rate and number of sweeps ....................................................................................... 44
6.3.7 Interpreting results ................................................................................................................. 45
7
FURTHER DEVELOPMENTS ................................................................................................................ 46
ANNEX 1 – DISCUSSION OF A TRACTABLE APPROACH TO DEFINING AND MEASURING IBOC
SIGNALS AGAINST RF MASKS ............................................................................................................... 47
1. INTRODUCTION .............................................................................................................................. 49
1.1.
GOALS OF THIS PAPER................................................................................................................. 50
2. TRACTABLE AND SUGGESTED OPERATIONAL SPECIFICATIONS FOR HYBRID IBOC
SPECTRAL OCCUPANCY MEASUREMENTS .................................................................................. 51
2.1. HYBRID FM DIGITAL SIGNAL FLATNESS AND OUT-OF BAND EMISSIONS: SPECIFICATION 1—
TRACTABLE DEFINITION .......................................................................................................................... 51
2.2. OPERATIONAL VARIATIONS: BANDWIDTH ...................................................................................... 52
2.3. OPERATIONAL VARIATIONS: REFERENCE LEVEL ........................................................................... 53
2.4. OPERATIONAL VARIATIONS: AVERAGING TIME .............................................................................. 54
2.5. MEASURING HYBRID FM IBOC SPECTRAL OCCUPANCY AGAINST THE MASK: ................................ 55
2.6. MEASURING HYBRID FM DIGITAL SIGNAL SUBCARRIER GROUP POWER (I.E. PRIMARY MAIN, PRIMARY
EXTENDED UPPER OR LOWER SIDEBANDS): .............................................................................................. 56
2.7. FURTHER DISCUSSION OF FM IBOC MEASUREMENTS WITH SPECTRUM ANALYZERS ...................... 58
3. AM .................................................................................................................................................... 60
3.1. HYBRID AM DIGITAL SIGNAL FLATNESS AND OUT-OF-BAND EMISSIONS: SPECIFICATION 2 –
TRACTABLE DEFINITION .......................................................................................................................... 60
3.2. OPERATIONAL VARIATIONS: BANDWIDTH ...................................................................................... 61
3.3. OPERATIONAL VARIATIONS: REFERENCE LEVEL ........................................................................... 62
3.4. OPERATIONAL VARIATIONS: AVERAGING TIME .............................................................................. 62
3.5. MEASURING HYBRID AM IBOC SPECTRAL OCCUPANCY AGAINST THE MASK: ............................... 63
3.6. MEASURING HYBRID AM DIGITAL SIGNAL SUBCARRIER GROUP POWER (I.E. PRIMARY, SECONDARY,
AND TERTIARY UPPER OR LOWER SIDEBANDS): ........................................................................................ 65
3.7. FURTHER DISCUSSION OF AM IBOC MEASUREMENTS WITH SPECTRUM ANALYZERS ..................... 66
4. DISCUSSION OF SPECTRUM ANALYZER MEASUREMENTS ................................................... 67
4.1. MAKING MEASUREMENTS TO THE STANDARDS WITH NEWER SPECTRUM ANALYZERS ..................... 67
4.2. MAKING MEASUREMENTS TO THE STANDARDS WITH OLDER SPECTRUM ANALYZERS: EXACT BUT
INCONVENIENT TECHNIQUES.................................................................................................................... 67
4.3. MAKING MEASUREMENTS TO THE STANDARDS WITH OLDER SPECTRUM ANALYZERS: INEXACT BUT
CONVENIENT TECHNIQUE ........................................................................................................................ 68
3.4.1.
How Close to Gaussian is the IBOC Digital Signal? ................................................... 68
4.4. AVERAGING TECHNIQUES ............................................................................................................. 69
4.4.1.
Trace Averaging vs. Video Filtering .............................................................................. 69
4.4.2.
Long Sweep Versus Trace Averaging........................................................................... 69
ANNEX 2: TEST AND MEASUREMENT EQUIPMENT SELF-CERTIFICATION LIST ............................ 71
Page 4
NRSC-G201-B
ANNEX 3: TEST AND MEASUREMENT EQUIPMENT SELF-CERTIFICATION FORM ......................... 73
ANNEX 4 – A METHOD FOR MEASURING HYBRID FM IBOC SIGNALS ON TRANSMISSION
SYSTEMS WITH INDEPENDENT DIGITAL AND ANALOG TRANSMISSION LINES USING A CHIMP
(COMBINED HYBRID IBOC MEASUREMENT PACKAGE) ..................................................................... 76
ANNEX 5 – RECOMMENDED AM ANTENNA BANDWIDTH CHARACTERISTICS ............................... 90
ANNEX 6 – AM TRANSMITTER MODULATION TECHNICAL PRIMER FOR NRSC MEASUREMENT
GUIDELINE ................................................................................................................................................. 96
Page 5
NRSC-G201-B
FIGURES
FIGURE 1. NRSC-5-C HYBRID FM IBOC WAVEFORM NOISE AND EMISSION LIMITS ............................................ 12
FIGURE 2. HIGH-POWER COMBINERS SUCH AS THE ONE SHOWN HERE ARE USED TO COMBINE THE OUTPUTS OF
SEPARATE ANALOG AND DIGITAL TRANSMITTERS AND REQUIRE A DUMMY LOAD WHICH WILL TYPICALLY
DISSIPATE A SIGNIFICANT AMOUNT OF POWER. (COURTESY OF SHIVELY LABS) ......................................... 14
FIGURE 3. EXAMPLE OF AN INTERLEAVED ANTENNA WHICH IS USED TO “SPACE COMBINE” THE ANALOG AND
DIGITAL PORTIONS OF AN FM IBOC SIGNAL. NOTE THAT EVERY OTHER BAY IS FED THE ANALOG SIGNAL
DIRECTLY BY THE HIGH POWER TRANSMISSION HARD-LINE RUNNING VERTICALLY BEHIND THE BAYS. THE
ALTERNATE BAYS ARE FED THE DIGITAL SIGNAL COMPONENT BY THE BLACK COAXIAL CABLES RUNNING
HORIZONTALLY FROM THE TOWER TO THE ARRAY (DENOTED BY YELLOW ARROWS). CLOSE INSPECTION OF
THE ANTENNA ELEMENTS REVEALS THAT THE ALTERNATING BAYS ARE CIRCULARLY POLARIZED IN OPPOSITE
DIRECTIONS. (COURTESY OF SHIVELY LABS) .......................................................................................... 15
FIGURE 4. TWO EXAMPLES OF SIMPLE COMMON-LINE FM SYSTEMS .................................................................. 16
FIGURE 5. EXAMPLE OF SPURIOUS EMISSIONS SAMPLED FROM THE OUTPUT OF AN FM IBOC TRANSMITTER. THE
SPURIOUS EMISSIONS IN THIS IMAGE ARE GENERATED BY THE TRANSMITTER AND ARE PREDOMINANTLY THE
RESULT OF SOME UNWANTED SIGNALS OVERLOADING THE TRANSMITTER’S POWER AMPLIFIER. SEE FIGURE 7
FOR AN EXAMPLE OF IBOC SPECTRAL REGROWTH AT 164 KHZ INTERVALS. (COURTESY OF BROADCAST
SIGNAL LAB) ......................................................................................................................................... 17
FIGURE 6. TWO EXAMPLES OF SIMPLE SEPARATE-LINE FM SYSTEMS ................................................................ 18
FIGURE 7. SOME OF THE COMMON DISTORTIONS ENCOUNTERED IN FM IBOC SIGNALS INCLUDE “WIDENED
SHOULDERS” OF THE DIGITAL OFDM SUBCARRIER SIDEBANDS AND “SPECTRAL REGROWTH” WHICH APPEARS
BEYOND THE MAIN PORTION OF THE SIGNAL. (COURTESY OF BROADCAST SIGNAL LAB) ............................ 21
FIGURE 8. NRSC-5-C HYBRID AM IBOC WAVEFORM SPECTRAL EMISSIONS LIMITS FOR 5 KHZ MODE ANALOG
BANDWIDTH .......................................................................................................................................... 25
FIGURE 9. NRSC-5-C HYBRID AM IBOC WAVEFORM SPECTRAL EMISSIONS LIMITS FOR 8 KHZ MODE ANALOG
BANDWIDTH .......................................................................................................................................... 26
FIGURE 10. AM FIELD MEASUREMENTS ARE BEST MADE WITH A GOOD QUALITY SHIELDED LOOP ANTENNA.
(COURTESY OF CHRIS SCOTT & ASSOCIATES) ....................................................................................... 28
FIGURE 11. AM IBOC SPECTRAL REGROWTH CAN BE CAUSED BY NON-LINEARITIES IN THE AMPLIFICATION
SYSTEM, BIAS ERRORS, AND REFLECTIONS CAUSED BY MISMATCHES BETWEEN THE TRANSMITTER AND
ANTENNA SYSTEMS. (COURTESY OF BROADCAST SIGNAL LAB) ................................................................ 30
FIGURE 12. LIMIT LINES ARE A POWERFUL TOOL AND CAN GREATLY SIMPLIFY MASK COMPLIANCE MEASUREMENTS.
(COURTESY BURT W EINER ASSOCIATES AND BROADCAST SIGNAL LAB) .................................................. 38
FIGURE 13. W HEN A MODULATED SIGNAL IS USED TO SET THE 0 DBC REFERENCE LEVEL, USE A WIDE RESOLUTION
BANDWIDTH SETTING (300 KHZ IS USED HERE), THEN USE THE POWER LEVEL AT THE CENTER FREQUENCY
TO ESTABLISH THE REFERENCE LEVEL. (COURTESY OF BROADCAST SIGNAL LAB) .................................... 40
FIGURE 14. W HEN A MODULATED AM SIGNAL IS USED TO SET THE 0 DBC REFERENCE LEVEL, AVERAGE THE
ANALYZER DISPLAY OVER MANY TRACES, THEN USE THE POWER LEVEL AT THE CENTER FREQUENCY TO
ESTABLISH THE REFERENCE LEVEL. (COURTESY OF BROADCAST SIGNAL LAB) ......................................... 44
TABLES
TABLE 1. NRSC-5-C HYBRID FM IBOC WAVEFORM NOISE AND EMISSION LIMITS.............................................. 12
TABLE 2. NRSC-5-C HYBRID AM IBOC WAVEFORM SPECTRAL EMISSIONS LIMITS FOR 5 KHZ MODE ANALOG
BANDWIDTH .......................................................................................................................................... 25
TABLE 3. NRSC-5-C HYBRID AM IBOC WAVEFORM SPECTRAL EMISSIONS LIMITS FOR 8 KHZ MODE ANALOG
BANDWIDTH .......................................................................................................................................... 26
Page 6
NRSC-G201-B
NRSC-5 RF MASK COMPLIANCE: MEASUREMENT METHODS
AND PRACTICE
1
SCOPE
This is an informative Guideline document which sets forth recommended methods and practices for
determining if hybrid FM and AM IBOC digital radio transmissions fall within the RF masks specified in the
NRSC-5 Standard. While not dealt with explicitly, the information on hybrid FM IBOC signals applies to
extended hybrid FM IBOC signals, as well. The NRSC anticipates incorporating information pertaining to
all-digital IBOC transmissions (as specified by NRSC-5) at the time when such transmissions are
authorized by the Federal Communications Commission (FCC).
2
2.1
REFERENCES
Normative References
This is an informative specification. There are no normative references.
2.2
Informative References
The following references contain information that may be useful to those implementing this Guideline
document. At the time of publication the editions indicated were valid. All standards are subject to
revision, and users of this Guideline document are encouraged to investigate the possibility of applying
the most recent editions of the standards listed below.
[1]
NRSC-5-C In-band/on-channel Digital Radio Broadcasting Standard, National Radio Systems
Committee, September 2011
[2]
NRSC-2-A Emission Limitation for Analog AM Broadcast Transmission, National Radio Systems
Committee, September 2007
[3]
Code of Federal Regulations (CFR) 47, Part 73, Subpart C – Digital Audio Broadcasting, Office of
the Federal Register, National Archives and Records Administration
[4]
HD Radio Implementation, Thomas R. Ray, Focal Press, 2008
[5]
The IBOC Handbook: Understanding HD Radio Technology, David P. Maxson, Focal Press, 2007
[6]
The NAB Engineering Handbook, 10th Edition, Edmund A. Williams, Editor-in-chief, Focal Press,
2007
[7]
The Role of the Detector in Spectrum Analyzer Measurement of Hybrid Digital Signals, David P.
Maxson, pp. 397-404, 2008 NAB Broadcast Engineering Conference Proceedings
[8]
Isolation in FM IBOC Multi-Channel Systems-St. Louis, Real World Data, Henry Downs, pp. 41-46,
2007 NAB Broadcast Engineering Conference Proceedings
[9]
Rebuilding a Legend: Rebuilding WOR Radio from the Tip of the Mic to the Top of the Tower,
Thomas R. Ray, pp. 243-264, 2007 NAB Broadcast Engineering Conference Proceedings
Page 7
NRSC-G201-B
[10]
Beyond Spectral Occupancy: an Investigation of IBOC Signal Quality Metrics, David P. Maxson,
pp.345-352, 2006 NAB Broadcast Engineering Conference Proceedings
[11]
Transmitter Output Power Measurements in Digital Broadcast Systems, Tim Holt, pp. 361-366,
2006 NAB Broadcast Engineering Conference Proceedings
[12]
Improvements to FM and IBOC Signal Quality through the Use of Pre-equalization, Mike Woods,
pp. 391-396, 2006 NAB Broadcast Engineering Conference Proceedings
[13]
8 Hints for Better Spectrum Analysis, Application Note 1286-1, Agilent Technologies March 27,
2008
[14]
Spectra and Bandwidth of Emissions, ITU-R Recommendation SM.328-11 (05/06)
[15]
AM IBOC Ascertainment Project, NAB Broadcast Engineering Conference, slides,
CPB_IBOC_NAB_2005.pdf, duTreil, Lundin & Rackley/Hatfield & Dawson joint venture
[16]
Evaluation and Improvement of AM Antenna Characteristics for Optimal Digital Performance,
Ronald D. Rackley, pp. 206-214, 2004 NAB Broadcasting Engineering Conference Proceedings
[17]
Spectrum Analyzer Measurements and Noise, Agilent Technologies, Inc., Application Note 1303,
April 2, 2008
[18]
AM/FM IBOC Measurements with the Agilent N9340B Handheld Spectrum Analyzer, Agilent
Technologies, Inc., Application Note 5989-9969EN, November 14, 2008
[19]
Pre-correction of Analog FM signals to correct for filterplexer distortions, Anders Mattsson, pp.397401, 2006 NAB Broadcast Engineering Conference Proceedings
[20]
Transmission Signal Quality Metrics for FM IBOC Signals, iBiquity Digital Corporation, Reference
Document SY_TN_2646s, Rev. 02, August 24, 2011
2.3
Symbols and abbreviations
In this Guideline the following abbreviations are used:
AM
ATU
BER
DANL
EPM
ERP
FCC
FM
IBOC
IMD
MER
MF
OFDM
PA
PSD
QAM
Amplitude Modulation
Antenna Tuning Unit
Bit Error Ratio
Displayed Average Noise Level
Equipment Performance Measurements
Effective Radiated Power
Federal Communications Commission (U.S.)
Frequency Modulation
In-Band/On-Channel
Intermodulation distortion
Modulation Error Ratio
Medium Frequency
Orthogonal Frequency Division Multiplexing
Power Amplifier
Power Spectral Density
Quadrature Amplitude Modulation
Page 8
NRSC-G201-B
QPSK
RBW
RF
S/N
VBW
VHF
2.4
Quadrature Phase Shift Keying
Resolution Bandwidth
Radio frequency
Signal-to-Noise ratio
Video Bandwidth
Very High Frequency
Definitions
In this Guideline the following definitions are used:
All digital waveform
A transmitted waveform for modes that do not include the analogmodulated signal. For FM IBOC, the all-digital waveform is composed
entirely of digitally modulated subcarriers, while for AM IBOC, the alldigital waveform is composed of digitally modulated subcarriers and the
unmodulated AM carrier.
Extended hybrid waveform
A transmitted waveform for modes composed of the analog FM signal
plus digitally modulated primary main subcarriers and some or all primary
extended subcarriers. This waveform will normally be used by
broadcasters requiring additional digital capacity over that provided by
the hybrid mode of operation (provides up to approximately 50 kbps
additional capacity).
HD Radio™
Trademark (of DTS, Inc.) for the digital AM and digital FM transmission
technology authorized by the FCC. Note that in the NRSC-5 Standard
and its normative references, the use of the term “HD Radio” is
interpreted as the generic term “IBOC” and should not be construed as a
requirement to adhere to undisclosed private specifications that are
required to license the HD Radio name from its owner.
Hybrid waveform
A transmitted waveform for modes composed of the analog -modulated
signal, plus digitally modulated primary main subcarriers. This waveform
will normally be used during an initial transitional phase preceding
conversion to the all-digital waveform.
Impedance bandwidth
The complex-impedance-versus-frequency characteristics of a system
such as an AM transmission system. The complex impedance is
commonly represented directly or indirectly in various ways, such as
phase/amplitude, VSWR, and real/imaginary mathematical notation.
OFDM subcarrier
A narrowband PSK or QAM-modulated carrier within the allocated
channel, which, taken together with all OFDM subcarriers, constitute the
frequency domain representation of one OFDM symbol.
Pattern bandwidth
A description of the variations in RF frequency response versus azimuth
of an AM antenna pattern.
Page 9
NRSC-G201-B
Resolution bandwidth (RBW) In a spectrum analyzer, the nominal bandwidth of the power detecting
element of the device. The RBW is adjustable and often must be set at a
specific value when trying to establish compliance with an RF mask.
RF mask
The graphical representation of the allowable RF signal power spectral
density (relative to a specific bandwidth) versus frequency for an RF
transmission. Typically, the power values are indicated relative to the
power of an unmodulated signal at the center frequency of the signal.
Spectral regrowth
The phenomenon whereby the signal energy of a band-limited RF signal
increases outside of the signal’s necessary bandwidth, typically due to
the existence of nonlinearities in the RF system.
Spectrum analyzer
An instrument used to characterize the amplitude (power) versus
frequency characteristics of a signal.
Trap filter
An RF filter used to strongly attenuate the energy within a certain range
of frequencies, typically characterized by a deep notch at the trapped
frequencies and low insertion loss at the desired frequencies.
Page 10
NRSC-G201-B
3
BACKGROUND
The purpose of an RF mask is twofold – interference control, and quality control. For controlling
interference, the RF mask establishes the limits for the unwanted emissions (spurious and out of band
1
emissions) and the limits for the desired emissions (within the necessary bandwidth). Adherence to the
mask improves the utilization of the spectrum and reduces interference potential among signals sharing
the spectrum. By implementing an RF mask, particularly in the context of digital signals, intermodulation
products, which are indicators of distortions of the transmitted signal, are kept within specified limits. The
RF mask therefore also acts as a basic control on the quality of the transmitted signal.
IBOC transmission architectures have evolved to adapt to the engineering requirements of various types
of broadcast facilities. Initially, FM IBOC facilities employed high-level combining to inject a digital signal
into the transmission line of the analog signal destined for the antenna.
Various FM IBOC
implementations now combine the analog and digital signals at each of many points in the transmission
chain, from the signal generation end (within the exciter) to the radiation end (at the antenna or in free
space). Each architecture places constraints on the performance of the transmission equipment and on
the methods of measuring the combined hybrid FM IBOC signal.
AM IBOC transmission architecture is not varied to the extent that it is with FM IBOC. However, AM
IBOC systems have site-specific constraints largely dependent on the design and type of antenna system
employed by the AM station. The nature of the relatively narrow bandwidth AM transmission system
presents a challenge to the engineer attempting to measure conformance with the RF mask. What goes
into the antenna system is not necessarily what comes out. Field measurement of AM signals is also
potentially challenging because of the presence of potentially strong environmental noise and
interference, including the energy of other stations within and near the necessary bandwidth of the AM
IBOC signal.
Mask compliance is by definition analyzed by evaluating the power spectral density (PSD) of the signal
and spurious emissions (PSD is described in the units of power per unit bandwidth, for example,
dBm/kHz). The most common instrument for performing such analysis is the swept spectrum analyzer.
The bandwidth is established by the resolution bandwidth filter, which approximates the ideal bandwidth
employed to define the RF mask. Newer spectrum analyzers employ various digital techniques to
improve the quality and accuracy of the measurement; in some instances an instrument that appears to
be a swept spectrum analyzer is a fully digital RF analyzer that presents its results in a manner that looks
like a swept analyzer. It is expected that as new products are developed for the industry innovative
approaches to evaluating IBOC signals may evolve.
This Guideline is intended to support all hybrid and extended hybrid IBOC spectrum conformance work,
but focuses primarily on the hybrid IBOC signal as used in the broadcast facility. While much of the
discussion relates to where and how to make measurements at the transmitter site, this Guideline also
provides developers of individual system components with a consistent and repeatable method of
evaluating those components when they are deployed at the customer’s broadcast facility. Such
consistency of measurement among manufacturers fosters both competition and interoperability.
1
Reference [14] defines occupied bandwidth and necessary bandwidth of the desired emissions, as well as
unwanted emissions, consisting of out-of-band emissions and spurious emissions.
Page 11
NRSC-G201-B
4
HYBRID FM IBOC mask compliance measurements
4.1
Overview
For hybrid FM IBOC, the NRSC-5 Standard (reference [1]) specifies, in Section 4.2.8, limits for noise and
spuriously generated signals from all sources, including phase noise and intermodulation products. This
specification is included here in Figure 1 and Table 1 and associated text and references from [1].
Figure 1. NRSC-5-C hybrid FM IBOC waveform noise and emission limits
NOTE: the upper and lower sidebands may differ in average power level by up to 10 dB
(asymmetric sidebands). Normally, the sideband power levels are equal, but under
certain scenarios, asymmetric sidebands may be useful for mitigation of adjacent channel
interference. Figure 1 shows a power-level difference of 10 dB for purposes of
illustration. It shall be noted that even though the upper and lower sidebands have
different power levels, the upper and lower spectral emissions limits are the same.
Table 1. NRSC-5-C hybrid FM IBOC waveform noise and emission limits
Frequency
offset relative
to carrier, kHz
100 - 200
200 - 250
250 - 540
540 - 600
> 600
2
Level relative to unmodulated carrier,
dBc/kHz
-40
-61.4 – (|frequency in kHz| - 200) x 0.260
-74.4
-74.4 – (|frequency in kHz| - 540) x 0.293
-80
Table 1 may be found in reference [6] (SY_SSS_1026s rev. F) of [1] (NRSC-5-C), Table 4-1.
Page 12
2
NRSC-G201-B
4.2
Modulating signals and compliance measurements
The signal(s) used for modulating the analog portion of the HD Radio hybrid IBOC signal can introduce
variance in RF mask compliance measurements. The shape of the PSD envelope of the analog portion
3
of the hybrid signal is determined by the modulating signals applied to the analog carrier. To determine
if a transmitter is performing to manufacturer specifications for RF mask compliance, the recommended
test configuration is the following:
●
●
●
●
Terminate the transmitter output in a dummy load of adequate power rating;
Operate the transmitter at the power level specified for the broadcast facility;
Apply the recommended test signal:
- Remove all modulation from the analog FM signal (no stereo pilot, no stereo subcarrier or
other FM subcarriers)
- Apply only a 1 kHz monaural tone at +/- 75 kHz total FM carrier deviation
Simultaneously transmit the HD Radio sidebands.
The recommended test signal is repeatable, thereby eliminating variances in results due to the
modulation. The recommended test signal has been found to generate fairly pronounced FM sideband
components, which then contribute to the generation of more pronounced 3rd- and 5th-order
intermodulation products if non-linearities are present in the transmitter. The recommended test signal
has been found to be more “demanding” against the mask (by about 2 dB) than using “pink” noise or
4
program audio.
Transmitter setup with this test tone will result in more conservative operation under the mask when
modulating with typical program material. This test signal is recommended for transmitter compliance
testing into a dummy load. Regular program material may be used for final equipment performance
measurements when the transmitter is connected to the antenna system.
The analog FM modulation frequency of 1 kHz is recommended so that the FM deviation of the analog
carrier is not affected significantly by the presence or lack of audio pre-emphasis. Therefore, the 1kHz
monaural tone can be input via either a balanced audio input or the wideband, composite, input of an FM
exciter.
Operational mask measurements, with the transmitter connected to its transmission system, may be
made with the station’s normal program modulation.
4.3
Operational facilities
There are two types of hybrid FM IBOC facilities of concern to the individual who must make
measurements. The first is the “common-line” facility—the facility that combines the analog and digital
components of the hybrid FM signal at some point before it reaches the antenna. The composite hybrid
IBOC signal can be sampled from the common transmission line after all combining and filtering
3
Note that the shape of the PSD envelope of HD Radio digital sidebands is, on the average, independent of the
information being transmitted. This is because the modulation of the digital signals is maintained by processes that
randomize the modulating information in a white noise-like fashion.
4
Several modulation types were considered for RF mask compliance measurements, and the 1 kHz monaural tone
was selected as the most effective. It is relatively unaffected by pre-emphasis and produces simple repetitive
sidebands. In contrast, experiments have shown that it is not as effective to apply more complex modulation to the
analog FM carrier. For instance, modulation with a 15 kHz tone at +/- 75 kHz deviation with L = -R (stereo mode),
and unmodulated subcarriers at 67 and 92 kHz (both at 5% injection), results in substantial analog spectrum
spreading, which contributes to less pronounced 3rd and 5th order RF intermodulation distortion (IMD) products.
Page 13
NRSC-G201-B
components in the system. The second is the “separate-line” facility where the composite hybrid signal is
not available for sampling on any transmission line.
Common-line facilities include those that use high power combining (Figure 2), split-level combining, lowlevel combining, or direct synthesis of the hybrid signal. Measurement of common-line facilities is readily
done with a single transmission line tap.
Figure 2. High-power combiners such as the one shown here are used to combine the outputs of
separate analog and digital transmitters and require a dummy load which will typically dissipate a
significant amount of power. (Courtesy of Shively Labs)
Separate-line facilities never combine the digital and analog signals within the transmission plant. In each
case of the separate-line facility, there are two transmission lines extending between the IBOC station’s
transmitters and antennas. Measurement of separate-line facilities requires careful calculation of the
individual gains and losses on the digital and analog systems to establish a reference level between
them.
Separate-line facilities include those that use separate antennas or dual-fed antennas. Separate antenna
systems utilize two physically separated antennas, either on the same tower or on different towers within
relatively close proximity. Interleaved antenna systems employ two antenna arrays whose bays are
interleaved within the same aperture on the tower (Figure 3). Dual-fed antennas are antennas that have
two inputs per bay, linked to the antenna elements by a four-port hybrid combiner.
Page 14
NRSC-G201-B
Figure 3. Example of an interleaved antenna which is used to “space combine” the analog and
digital portions of an FM IBOC signal. Note that every other bay is fed the analog signal directly
by the high power transmission hard-line running vertically behind the bays. The alternate bays
are fed the digital signal component by the black coaxial cables running horizontally from the
tower to the array (denoted by yellow arrows). Close inspection of the antenna elements reveals
that the alternating bays are circularly polarized in opposite directions. (Courtesy of Shively Labs)
4.3.1
4.3.1.1
Common-line facilities
Measurement location and sampling method
To obtain a sample of the hybrid FM IBOC signal at a common-line broadcast facility, identify the last
component in the signal chain before the signal is sent to the antenna.
Figure 4 contains two examples of simple common-line FM systems. The stars mark the location at or
beyond which the analog and digital signals are on a common transmission line. If there are filters or a
master antenna combiner, obtain the RF sample after the last component. This ensures that the
measurement will capture any bandpass characteristics affecting signal flatness, and any possible nonlinear effects from other transmitters or arcing components in the transmission system.
Page 15
NRSC-G201-B
Digital
Gen
Digital
Gen
Σ
Analog
Gen
PA
Σ
PA
Analog
Gen
LOW-LEVEL COMBINED
PA
HIGH-LEVEL COMBINED
Figure 4. Two examples of simple common-line FM systems
To be certain that the common-line sample is representative of the radiated signal, two primary concerns
should be addressed—antenna bandwidth and sample quality. First, while most FM antennas are
sufficiently broadband to reliably radiate the hybrid IBOC signal, it is important to be certain that the
system being tested has an antenna that meets these requirements. Normally, knowing the frequency for
which the antenna was tuned, and the manufacturer’s specifications for the antenna should be sufficient.
In special cases such as those that involve a station at the edge of a master antenna’s passband, or an
antenna that may have been obtained inexpensively from a station on a different frequency, it is advisable
to sweep the antenna to determine if there are any significant bandwidth anomalies that require attention.
The second concern, and the one more likely to be a source of difficulty in obtaining a representative
sample of a hybrid FM IBOC signal, is the quality of the directional coupler employed to obtain a sample.
First of all, a simple signal sampling loop may not be reliable because it is not directional. Unwanted
energy feeding into the reverse path on the transmission line might affect the quality of the measurement
from the loop. Also, when employing a directional coupler, the coupler should have a directivity of at least
30 dB, and potentially more in challenging environments with strong reverse-path signals.
The RF sample should be strong enough to maximize the signal-to-noise ratio (S/N) in the measuring
instrument. At the same time, the level at which the RF sample arrives at the input to the measurement
instrument’s first mixer, converter or detector should not overload the instrument. This is discussed
further in Section 6.1.1.
4.3.1.2
Instrument configuration
When a reliable signal is obtained from the sample port, it is time to establish a power reference, set up
the instrument span and bandwidth, and configure the detector, sweep and averaging functions. For
further information see Section 6.2 on measuring.
4.3.1.3
Spurious emissions
The intermodulation of the digital and analog components of the hybrid FM IBOC signal potentially
produces spurious emissions at regular intervals (nominally 164 kHz on center) above and below carrier
frequency (see Figure 5). Further intermodulation products may occur in the manner that traditional
analog facilities sometimes experience at multi-transmitter sites. Just as the two analog signals of colocated stations may intermodulate and produce sum and difference products within and near the FM
band, so, too, may the analog and digital components of a hybrid signal interact with the analog and
Page 16
NRSC-G201-B
digital components of co-located stations’ signals. It is recommended that a third-order intermodulation
study be prepared before evaluating a new or modified facility. This may help the evaluator to identify
emissions observed during the test and to determine whether the station under test is the cause of the
spurious emission.
Figure 5. Example of spurious emissions sampled from the output of an FM IBOC transmitter.
The spurious emissions in this image are generated by the transmitter and are predominantly the
result of some unwanted signals overloading the transmitter’s power amplifier. See Figure 7 for an
example of IBOC spectral regrowth at 164 kHz intervals. (Courtesy of Broadcast Signal Lab)
While attention in this document is primarily focused on the desired emission and nearby out-of-band and
spurious emissions, transmitter harmonics are another critical source of spurious emissions. The hybrid
FM IBOC mask extends indefinitely outside the bounds of Figure 1, incorporating the station’s harmonic
frequencies. The sample point for the hybrid FM IBOC common-line signal should be after the harmonic
filter has had an opportunity to attenuate harmonic emissions.
The directional coupler may be broadband enough for establishing levels of spurious emissions within
and near the FM spectrum, depending on the passband of the directional coupler. However, additional
care must be taken with harmonic measurements. The coupler response may understate harmonic
power, thereby artificially lowering the apparent level of harmonic products. If the coupler’s response to
harmonics is known and a corresponding correction is made to the measured harmonic level, a reliable
worst-case harmonic measurement is obtained from the transmission line. It is reasonable to assume
that harmonics will be further attenuated by the antenna before being radiated.
When measuring harmonics, it is helpful to insert an FM band “trap filter” whose insertion loss and
response at the harmonic frequencies is known. This reduces the power of the incoming fundamental
frequency, reducing the possibility that the analyzer’s internal harmonic distortion will be mistaken for a
transmitted harmonic product.
Another valuable tip for working with harmonic and intermodulation products on an analyzer involves
testing an apparent spurious emission for whether it is emitted by the system under test or internally
generated by the measurement instrument. If the instrument is being forced into a non-linear range that
causes internal spurious products, the addition of attenuation to the input of the instrument will change
the spurious component by more than the value of the attenuator. For example, with the insertion of a 10
Page 17
NRSC-G201-B
dB attenuator, all external signals entering the instrument will be attenuated by 10 dB. However, any
internally generated spurious signals will diminish by more than 10 dB, because the internal process is
not linear. This is a helpful test when confronted with a spur and uncertain about the performance of the
instrument.
4.3.2
Separate-line facilities
4.3.2.1
Measurement location and sampling method
Obtaining a sample of the hybrid FM IBOC signal at a separate-line broadcast facility is more challenging
that at a common-line facility. There is no place, except over the air, from which to sample the hybrid FM
IBOC signal. Unfortunately, over-the-air measurements are subject to inconsistencies resulting from
differing antenna patterns for the analog and digital signals and to multipath in the signal received by the
5
test instrument. In addition, over-the-air reception is subject to noise, interference and signals on
channel adjacencies that could mask the presence of non-compliant emissions. Consequently, over-theair measurements are not recommended for assessing hybrid IBOC FM signals for RF mask compliance.
For separate-line system mask compliance measurements, it is recommended that independent analog
and digital signals samples be obtained and carefully compared to determine if mask compliance is
achieved. Figure 6 contains two examples of simple separate-line systems.
The stars mark the
locations at or beyond which the independent analog and digital signals should be sampled. If there are
filters or a master antenna combiner, obtain the RF sample after the last component. This ensures that
the measurement will capture any bandpass characteristics affecting signal flatness, and any possible
non-linear effects from other transmitters or arcing components in the transmission system.
Digital
Gen
PA
Analog
Gen
PA
Digital
Gen
PA
(NOTE – ANTENNAS MAY BE
ON THE SAME TOWER)
Analog
Gen
DUAL-FED ANTENNA
PA
SEPARATE ANTENNA
Figure 6. Two examples of simple separate-line FM systems
To be certain that the separate-line sample ports obtain samples that are representative of the radiated
signal, two primary concerns should be addressed—antenna bandwidth and sample quality. First, while
most FM antennas are sufficiently broadband to reliably radiate the hybrid IBOC signal, it is important to
be certain that the system being tested has an antenna that meets these requirements. Normally,
knowing the frequency for which the antenna was tuned, and the manufacturer’s specifications for the
antenna should be sufficient. In special cases that involve a station at the edge of a master antenna’s
passband, or an antenna that may have been obtained inexpensively from a station on a different
5
It is recommended practice to have the separate antenna systems for analog and digital signals operate with
antenna vertical and horizontal patterns, antenna heights and antenna positions that are as similar as possible to
minimize variations in the analog-to-digital power ratios in the field. See also §73.404 (d) of the FCC rules.
Page 18
NRSC-G201-B
frequency, it is advisable to sweep the antenna to determine if there are any significant impedance
bandwidth anomalies that require attention.
The second concern, and the one more likely to be a source of difficulty in obtaining a representative
sample of a hybrid FM IBOC signal is the quality of the directional coupler employed to obtain a sample.
First of all, a simple signal sampling loop may not be reliable because it is not directional. Unwanted
energy feeding into the reverse path on the transmission line might affect the quality of the measurement
from the loop. Also, when employing a directional coupler, the coupler should have a directivity of at
least 30 dB, and potentially more in challenging environments with strong reverse-path signals.
This is a particular concern in separate-line systems because the isolation between the digital and analog
transmission chains varies with the facility design. Low isolation figures can place greater demands on
coupler directivity to avoid false indications of spurious emissions. In addition, the isolation in the
separate-line systems may be insufficient for protecting the digital and the analog transmitters from
ingesting too much energy from the other transmitter. This is especially the case with a digital transmitter
that is confronted with an incoming (reverse path) analog signal with substantial power. Either
transmitter, but especially the digital transmitter, may produce spurious emissions and digital signal
distortions that could be masked by poor coupler directivity. Extra attention to isolation and coupler
performance should be paid in the design of a separate-line hybrid FM IBOC system.
Since there is no place in the separate-line system where the ratio of the analog and digital signals is
established empirically, special attention must be given to the system gains and losses on both the
analog and digital systems. The ratio between the power of the digital and analog signals must be set by
calculation and system design because they cannot be reliably observed in separate-line systems. The
signal samples taken from directional couplers on the separate analog and digital transmission lines must
be adjusted in power to obtain the correct relative power levels. The differences in line losses and
antenna gains between the sample points and the respective antennas must be accounted for. In
addition, the couplers’ coupling ratios should be known as precisely as possible. With this information,
one can make a reasonable comparison between the spectrum samples taken from the analog and digital
lines and the RF mask, and in doing so, determine the analog to digital power ratio.
A tool to simplify the comparison of the analog and digital signals on separate-line systems has been
studied. It is a test jig that combines the analog and digital signal samples from the separate lines to
create a virtual common-line copy of the combined signals. It has been called the Combined Hybrid
IBOC Measurement Package (“CHIMP”). In Annex 4, a white paper on the device is presented explaining
how to make and use such a device and how the performance of the device was validated.
When performing separate-line sampling, the RF sample should be strong enough to maximize the S/N in
the measuring instrument. At the same time, the level at which the RF sample arrives at the input to the
measurement instrument’s first mixer, converter or detector should not overload the instrument. This is
discussed further in Section 6.1.1.
4.3.2.2
Instrument configuration
When reliable signals are obtained from the sample ports, it is time to establish a power reference, set up
the instrument span and bandwidth, and configure the detector, sweep and averaging functions. For
further information, see Section 6.2 on measuring.
4.3.2.3
Spurious emissions
The intermodulation of the digital and analog components of the hybrid FM IBOC signal potentially
produces spurious emissions at regular intervals (nominally 164 kHz on center) above and below carrier
frequency. Further intermodulation products may occur in the manner that traditional analog facilities
Page 19
NRSC-G201-B
sometimes experience at multi-transmitter sites. Just as the two analog signals of co-located stations
may intermodulate and produce sum and difference products within and near the FM band, so too may
the analog and digital components of a hybrid signal interact with the analog and digital components of
co-located stations’ signals. It is recommended that a third-order intermodulation study be prepared
before evaluating a new or modified facility. This may help the evaluator to identify emissions observed
during the test and to determine whether the station under test is the cause of the spurious emission.
While attention in this document is primarily focused on the desired emission and nearby out of band and
spurious emissions, transmitter harmonics are another critical source of spurious emissions. The hybrid
FM IBOC mask extends indefinitely outside the bounds of Figure 1, incorporating the station’s harmonic
frequencies. The sample points for the hybrid FM IBOC separate-line signals should be after the
harmonic filters have had an opportunity to attenuate harmonic emissions.
The directional coupler used for signal sampling may be broadband enough for establishing levels of
spurious emissions within and near the FM spectrum, depending on the passband of the directional
coupler. However, additional care must be taken with harmonic measurements. The coupler response
may understate harmonic power, thereby artificially lowering the apparent level of harmonic products. If
the coupler’s response to harmonics is known and a corresponding correction is made to the measured
harmonic level, a reliable worst-case harmonic measurement is obtained from the transmission line. It is
reasonable to assume that harmonics will be further attenuated by the antenna before being radiated.
When measuring harmonics, it is helpful to insert an FM band trap filter whose insertion loss and
response at the harmonic frequencies is known. This reduces the power of the incoming fundamental
frequency, reducing the possibility that the analyzer’s internal harmonic distortion will be mistaken for a
transmitted harmonic product.
Another valuable tip for studying harmonic and intermodulation products on an analyzer involves testing
an apparent spurious emission for whether it is emitted by the system under test or internally generated
by the measurement instrument. If the instrument is being forced into a non-linear range (due to signal
overload) that causes internal spurious products, the addition of attenuation to the input of the instrument
will change the spurious component by more than the value of the attenuator. For example, with the
insertion of a 10 dB attenuator, all external signals entering the instrument will be attenuated by ten dB.
However, any internally generated spurious signals will diminish by more than 10 dB, because the
internal process is not linear. This is a helpful test when confronted with a spur and uncertain about the
performance of the instrument.
Also, for separate-line systems, both the analog transmission line and the digital transmission line may
show outgoing spurious energy on the same part of the spectrum. It is difficult to obtain a reliable sum of
the power density spectrum of each to establish total power in the radiated spurious signal. The sum of
co-frequency spurious emissions on the analog transmission line and the digital transmission line may to
be non-compliant with the RF mask, while the individual images appear compliant. If so, further work is
necessary.
For example, if the individual spurious emission on each transmission line were 1 dB below the mask,
then the sum of these two co-frequency emissions would be 2 dB over the limit (the sum of two equal
power signals represents a 3 dB increase in total power). To sum the co-frequency spurious emissions of
both the analog and digital transmissions, the power values in decibels should be converted to linear
values, summed, and reconverted to decibels. This would yield a composite value that is an
approximation of the sum of the energy on the digital line and the analog line at that frequency.
Alternatively, use the measurement combiner box (CHIMP) described in Annex 4 to sum the two sampled
signals at the correct ratio to evaluate spurious emissions.
4.4
Troubleshooting (various out-of-spec conditions and possible remedies)
Page 20
NRSC-G201-B
Distorted signals in FM systems are typically the result of amplification problems. In situations where
there are filters and combiners, one must make certain that these devices are not excessively rolling off
the frequency and phase response in the critical passband of the hybrid IBOC signal. Intermodulation
products, generally caused by amplification problems, result in widened “shoulders” of the digital OFDM
subcarrier sidebands and in “spectral regrowth” that appears as “beehive” bumps of energy out of the
station’s occupied bandwidth (see Figure 7). Intermodulation may also create the appearance of tilted
OFDM sideband levels.
Figure 7. Some of the common distortions encountered in FM IBOC signals include “widened
shoulders” of the digital OFDM subcarrier sidebands and “spectral regrowth” which appears
beyond the main portion of the signal. (Courtesy of Broadcast Signal Lab)
4.4.1
Pre-correction and clipping discussion
It is important to maintain amplifier bias levels that conform to factory specifications. If the bias is too
high, there is insufficient headroom for the power amplifier (PA) to handle the peaks. Also, if the exciter
has fixed pre-correction (in which there is a factory or user setting that is set and stays put), it should be
6
adjusted to anticipate the level of bias on the transmitter. Poorly amplified IBOC signals will have
considerable spectral regrowth and gain flatness may also appear to suffer.
4.4.2
Evaluating ingress of unwanted signals
Any system with two final PAs in parallel has the capacity to feed back energy from one into the other,
potentially producing intermodulation products in the PAs. At multiple-station sites, signal energy from
6
“Pre-correction” refers to the technique by which the phase and/or magnitude response of the signal generating
device is intentionally distorted to compensate for the overall phase and magnitude response of the system. See
reference [19] for additional information.
Page 21
NRSC-G201-B
other stations may also find its way into the output of the PA, complicating the situation with more
intermodulation products. This unwanted energy imposed on a PA output can be one of the primary
causes of non-complaint hybrid FM IBOC emissions.
On separate-line systems there is some degree of isolation between the two systems. The isolation can
be measured using the reverse direction of the transmission line directional coupler. This is best done
with one transmitter off-air while measuring the energy received by the transmission line of that off-air
transmitter. The coupler has associated with it a coupling loss (for example, 50 dB below the power of
the energy on the line in the direction being sampled) that can be used to help determine the isolation
between the systems as shown in Equation 1:
Equation 1 (decibel addition & subtraction)
(Measured level of incoming signal)
+ (coupling loss)
– (Tx1 power out)
= (isolation between Tx1 and the sample point of the Tx2 system)
Turnaround loss of a transmitter indicates how much of the unwanted energy coming into its PA from the
transmission line is turned around and sent back out to the antenna. Turnaround loss can be on the
order of 6 dB, depending on the transmitter. Knowing the turnaround loss can help in estimating how
much energy from one transmitter will be output by the other transmitter as shown in Equation 2:
Equation 2 (decibel addition & subtraction)
(Tx1 output power)
– (isolation)
– (turnaround loss of Tx2)
= (expected output power of Tx2’s turning around Tx1 energy)
In other words, (isolation) – (turnaround loss) indicates how many dB below the Tx1 transmitted signal will
be the distorted copy of the transmitted signal emitted by the Tx2 PA. To relate these to effective
radiated power (ERP), one must account for line losses and antenna gains.
On systems with common-line configurations, turnaround loss is of most concern when high-level and
split-level combining are used. Low-level combining does not present the same design challenges as
combining the outputs of high power amplifiers. Direct synthesis begins with a combined hybrid signal
from the digital-to-analog converter and has no active components that are separately handling analog
and digital waveforms.
High-level and split-level combining can be evaluated with directional couplers on the outputs of the
transmitters, before the combining network. High-level combining can be evaluated in the same manner
as described above for separate-line evaluation, that is, by shutting off one transmitter while evaluating
the isolation from the other transmitter. Since split-level combining requires the presence of some analog
energy on both transmitters to make the hybrid combiner function as planned, measurements made by
turning off one transmitter to measure the isolation to the other transmitter may be unreliable.
Care must be taken to ensure that any measurement in one direction on a directional coupler is not
compromised by signal traveling in the other direction on the transmission line. Directional couplers have
a “directivity” figure that indicates how much the forward signal crosstalks into the reverse port and vice
versa. Measurements can be limited by the directivity of a coupler, especially when the coupler has a
strong signal passing through it in the direction opposite to the direction being sampled.
Page 22
NRSC-G201-B
4.4.3
Digital Signal Quality Testing
Digital waveforms can be analyzed for their deviation from the corresponding ideal waveform . This is a
powerful technique to characterize the overall “health” of a transmitted signal and help troubleshoot
problems with transmitted IBOC signals. Distortions of the waveform increase the chance of receiving
errors in the symbols used to represent the digital information. These distortions can be measured and
the results can be compared to recommended specifications.
Timing, amplitude, and phase distortions encountered in the transmission plant, the RF channel, and the
receiver all contribute to errors in reception of the digital radio signal. Engineers try to minimize the
distortion in the transmitted signal in order to reserve an error margin for the distortions contributed by the
RF channel and the receiver. In other words, by transmitting as distortion free a signal as is reasonably
possible, the broadcaster maximizes the error correction “headroom” available at the receiver for the
impairments caused by the propagation and reception of the digital signal By making digital signal
quality measurements at the transmission plant, the amount of headroom available can be characterized.
Further, these measured characteristics can help identify problems in the transmission system in cases
where headroom is found to be inadequate.
4.4.3.1
Modulation Error Ratio (MER)
The method specified by iBiquity for performing digital signal quality testing on an FM IBOC signal is to
7
measure the Modulation Error Ratio (MER) as described in [20]. MER is a measurement of the signalto-noise ratio (SNR) of the binary information carried on the digital waveform. MER measurements tend
to give the broadcast engineer a more useful, “grayscale,” diagnostic view of system problems than the
more “black and white” view obtained from bit error ratio measurements (discussed in 4.4.3.2).
Reference [20] specifies three different MER measurements, plus group delay and amplitude response
derived from the reference carriers. The MER measurements include linear equalization, therefore they
are not degraded by moderate linear distortions. The MER measurements focus on nonlinear distortion,
noise, and PAR reduction noise. Linear distortions are quantified by separate measurements of amplitude
response and group delay. As discussed in [20], the specific manner in which MER is measured depends
8
upon which subcarriers—data subcarriers or reference subcarriers—are used for the measurement.
As with SNR, a higher MER indicates a better quality signal. As the MER decreases, the likelihood of
errors in reception increases. Satisfactory FM IBOC transmission quality is achieved when the average
MER of the Binary Phase Shift Keying (BPSK) reference subcarriers is sufficiently high for robust
reception.
Referring to the MER specification in [20], when measured at the RF output of the transmission system –
at the connection point to the antenna system after any RF filters or combining apparatus – the MER
averaged across all reference subcarriers shall be greater than or equal to 14 dB.
The MER and linear distortion measurement specification in [20] can be employed to measure the signal
quality of the overall digital transmission system including any RF filters and/or combiners. It is anticipated
that tools for making such measurements will become available for characterizing the quality of the digital
signal, including MER measurement capabilities within IBOC exciters as well as standalone MER
measurement devices (similar to analog modulation monitors). The NRSC may consider recommending
and/or adopting other specifications , including updates to the above referenced iBiquity documents on
signal quality metrics, as the IBOC industry continues to mature and new measurement tools and
techniques become available.
7
8
The “C’ version of the NRSC-5 Standard is expected to incorporate references to [20].
See [1] for additional information on the structure of an FM IBOC signal.
Page 23
NRSC-G201-B
4.4.3.2
Bit Error Ratio (BER)
In the digital domain, it is also customary to demodulate the digital waveform and evaluate the accuracy
of the resulting bit stream compared to the original. This can be reported as the bit error ratio (BER),
reporting the ratio of the number of bit errors to the total number of bits being evaluated. The “raw,”
uncorrected BER measures the demodulated data stream before error correction is applied in the
receiver. There is also a post-error correction BER that applies the digital system’s built-in error
correction features and identifies the bit errors that remain after error correction.
The post-error-correction BER in the receiver tends to exhibit a cliff effect, in that the output remains
relatively error-free over a wide range of input SNR, but then at some point the SNR reaches a
“threshold” after which a relatively small change in SNR overwhelms the system with errors and the
output error ratio then rises steeply, with abrupt system failure ensuing. For this reason, a post-error
correction BER will tend to mask subtle impairments in the signal under test, while the raw BER provides
more information about the number of errors occurring before the system fails.
4.4.3.3
MER vs. BER
There is somewhat of a paradox in the reception and characterization of received digital signals. The
measurement device must acquire the signal in order to measure it, but the act of acquiring relies on
aligning the measurement device to the signal under test. In addition to the error correction (which affects
the post-error correction BER), the digital signal receiver first must perform several steps to “lock on” to
the signal. These steps include frequency adjustment to center the receiver on the exact center
frequency of the waveform, timing adjustment to lock onto the symbols and identify the proper starting
point of each symbol, and phase equalization to adjust for the variation of the phase of the digital
waveform across the occupied channel. The pre- and post-error correction BER measurements rely on
the measurement device to perform these adjustments to acquire the signal.
Different receiver designs may behave differently under certain kinds of signal impairments and it is
therefore important for BER measurements to be performed with a uniform reference receiver design. By
contrast, MER measurements made at the transmitter may use simpler and more consistent algorithms
since there will be no interfering signals, Doppler shift, propagation effects, multipath, etc. If the MER
measurement is made at the IBOC exciter, the frequency and timing information is already, directly,
available along with both the original data and the demodulated data sample from the output of the
transmission system. Amplitude response and group delay measurements are derived from the same
OFDM demodulation which produces the MER measurements as described in [20]. These can measure,
among other aspects, the unequalized phase performance of the signal across the channel bandwidth.
MER measures linearly equalized signal quality whereas the BER relies on the receiver doing some work
to correct errors to the best of its ability. By taking an MER measurement at the transmission system
output to the antenna, the effects of the propagation channel and the effects of the receiver equalization
are removed from the measurement, revealing only the effects of the transmission system on the
waveform.
Page 24
NRSC-G201-B
5
HYBRID AM IBOC Mask Compliance measurements
5.1
Overview
For hybrid AM IBOC, the NRSC-5 Standard specifies, in Section 4.1.8, limits for noise and spuriously
generated signals from all sources, including phase noise and intermodulation products. This
specification is included here for 5 kHz mode (Figure 8 and Table 2) and 8 kHz mode (Figure 9 and Table
3).
0
dBc in a 300 Hz bandwidth
-10
-20
-30
-40
-50
-60
-70
-80
-90
-100
-110
-80
-75
-70
-65
-60
-55
-50
-45
-40
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
Frequency offset, kHz
Hybrid Spectral Emissions Limits / 5 kHz Analog Bandwidth
Nominal Digital Carrier Power Spectral Density
Nominal Analog Carrier Power Spectral Density / 5 kHz Analog Bandwidth
Figure 8. NRSC-5-C hybrid AM IBOC waveform spectral emissions limits for 5 kHz mode analog
bandwidth
Table 2. NRSC-5-C hybrid AM IBOC waveform spectral emissions limits for 5 kHz mode analog
bandwidth
Frequency
offset relative
to carrier, kHz
5 - 10
10 - 15
15 - 15.2
15.2 - 15.8
15.8 - 25
25 - 30.5
30.5 - 75
> 75
Level relative to unmodulated carrier,
dBc per 300 Hz
-34.3
-26.8
-28
-39 – (|frequency in kHz| - 15.2) x 43.3
-65
-65 – (|frequency in kHz| - 25) x 1.273
-72 – (|frequency in kHz| -30.5) x 0.292
-85
Page 25
80
NRSC-G201-B
0
dBc in a 300 Hz bandwidth
-10
-20
-30
-40
-50
-60
-70
-80
-90
-100
-110
-80
-75
-70
-65
-60
-55
-50
-45
-40
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
Frequency offset, kHz
Hybrid Spectral Emissions Limits / 8 kHz Analog Bandwidth
Nominal Digital Carrier Power Spectral Density
Nominal Analog Carrier Power Spectral Density / 8 kHz Analog Bandwidth
Figure 9. NRSC-5-C hybrid AM IBOC waveform spectral emissions limits for 8 kHz mode analog
bandwidth
Table 3. NRSC-5-C hybrid AM IBOC waveform spectral emissions limits for 8 kHz mode analog
bandwidth
Frequency
offset relative
to carrier, kHz
8 - 10
10 - 15
15 - 15.2
15.2 - 15.8
15.8 - 25
25 - 30.5
30.5 - 75
> 75
Level relative to unmodulated carrier,
dBc per 300 Hz
-34.3
-26.8
-28
-39 – (|frequency in kHz| - 15.2) x 43.3
-65
-65 – (|frequency in kHz| - 25) x 1.273
-72 – (|frequency in kHz| -30.5) x 0.292
-85
In addition to performing a measurement of the hybrid AM IBOC signal against the hybrid IBOC mask, the
NRSC-5-C standard also calls for a measurement of the analog-only AM signal against the FCC mask
described in 47 CFR §73.44.
5.2
Operational facilities
Field measurement of hybrid AM IBOC signals is necessary because the antenna system has such a
strong influence on the quality of the radiated signal that sampling the transmitter output may not be
representative of the signal characteristics. However, there are specific transmission sampling
techniques that make it possible to rely on transmitter output measurements to verify RF mask
compliance. Such verification is conditional and subject to field verification in the event that an
interference issue or mask compliance concern is raised. The specific sampling techniques are explained
below.
Page 26
NRSC-G201-B
The antenna system will for the most part not generate intermodulation-based spurious emissions on its
own. Only in the event of a significant nonlinearity in the antenna system (such as an arcing condition in
the mechanics of the system) is it possible that the antenna system will generate new intermodulation
products that might exacerbate spectral regrowth or create other spurious emissions. In addition to
transmitter amplification non-linearities, impedance mismatches between the transmitter and antenna
system can contribute to the creation of intermodulation products (spectral regrowth) in the transmitter
final amplifier. Furthermore, the impedance bandwidth and pattern bandwidth characteristics of the
antenna system can attenuate or enhance spectral regrowth sidebands with respect to the center
frequency and antenna azimuth.
Issues relating to the pattern bandwidth of the antenna system can affect the digital-to-analog power
ratios and can roll off certain portions of the hybrid signal at certain azimuths from the directional array.
This Guideline does not address pattern bandwidth issues, other than to recommend taking off-air
measurements within the main lobe of the pattern.
5.2.1
Directional and non-directional antenna systems
There is little difference between measuring directional and non-directional antenna systems. Directional
systems have a main lobe within which the hybrid IBOC signal should be at optimum level and
performance. Mask compliance in directional arrays, when measured in the field, is determined by
measurements made on a main lobe, within the ±3 dB geographic arc around the peak of the main lobe.
Non-directional systems may be measured in any direction from the antenna site where local noise and
interference are minimal and the desired signal strength is sufficient to provide good dynamic range in the
measurement.
5.2.2
5.2.2.1
Measurement locations and sampling methods
Field measurement
The reference location for measuring hybrid AM IBOC mask compliance in the field is the same as the
NRSC-2 measurement location. A location approximately 1 km from the antenna site is suggested. The
objective is to site the measuring equipment optimally:
a) in the far field of the antenna system;
b) as close as possible to the antenna to maximize signal level and to minimize measurement
noise and ambient noise; and
c) as remotely as possible from sources of interference, noise, and re-radiation that may
confound the results.
Measurements of stations with directional antenna patterns should be made at a point within ±3 dB of the
peak of the main lobe. For NRSC-5 mask compliance measurements, the reference level is the actual
analog signal power level observed at the time the measurement is taken.
A good quality shielded loop antenna is commonly employed to peak the signal and minimize electrical
noise pickup such as that shown in Figure 10.
Page 27
NRSC-G201-B
Figure 10. AM field measurements are best made with a good quality shielded loop antenna.
(Courtesy of Chris Scott & Associates)
5.2.2.2
True power measurement at transmitter
Since sometimes it can be difficult to obtain reliable measurements of AM IBOC mask compliance in the
field, a measurement from the AM transmission line using the proper technique can serve as an
acceptable secondary measurement method. However, if there is reason to believe that the transmission
line measurement is incorrect, a field measurement may be necessary. Transmission line samples may
be taken between the transmitter and the next component in the system, typically the common point or
antenna tuning unit (ATU) as the case may be.
The signal sampling technique used is critical to the accuracy of the measurement. Because the
impedance of a typical AM antenna system varies across the bandwidth of the hybrid IBOC signal, a
simple voltage or current sample will not suffice due to the uncertainty introduced by the voltage and
current standing wave in the presence of a load mismatch. Using a directional coupler eliminates the
error due to the standing wave and will provide accurate forward and reflected power samples. As a
result, the use of a directional coupler is recommended. Note that many installations and transmitters are
not equipped with directional couplers so this would need to be added in order to obtain accurate power
samples.
The sampling technique must obtain an accurate measurement of the real, delivered power accepted by
the antenna system across the frequency range of interest. Mathematically, this requires sampling the
voltage (magnitude and phase) and current (magnitude and phase) and forming the scalar product of the
two (“dot product”), taking into account forward and reflected power and vector summing the samples
(prior to making the dot product calculation).
Page 28
NRSC-G201-B
A practical method of obtaining a measurement of real, delivered power is to subtract the reflected PSD
from the forward PSD at each frequency of interest. Since the objective is to obtain the radiated PSD by
frequency, the power sampling method must be unaffected by the variation in impedance across the
bandwidth of interest.
Note that if the reflected PSD is not subtracted from the forward PSD the resulting measurement will be
higher than the real PSD and therefore will be a more conservative measurement of out-of-band
emissions. In general, this error is relatively small and may be acceptable without correction. For
example, if the reflected PSD measured is 10 dB below the forward PSD (return loss = 10 dB, VSWR =
1.9:1) the forward PSD measurement is 0.5 dB higher than the real PSD. If the reflected PSD measured
is 3 dB below the forward PSD (return loss = 3 dB, VSWR = 5.8) then the forward PSD is 1.8 dB higher
than the real PSD.
5.2.2.3
Impact of pattern bandwidth
Transmission line measurements do not account for the pattern bandwidth, which is a description of the
variations in frequency response versus azimuth of the AM antenna pattern. Typically, the performance
of the hybrid IBOC signal in the main lobe will be similar to the results of the transmitter output true power
sample method described above. On other radials, and especially at nulls, the center frequency may be
more depressed than other frequencies in the hybrid IBOC channel, or outside the channel. The result
can appear as distortion of the PSD of the hybrid IBOC signal compared to the main lobe and
transmission line measurements and the appearance of non-compliant OFDM subcarrier levels and
spectral regrowth levels with respect to the mask. For this reason, only field measurements in the main
lobe are applicable to compliance measurement.
5.2.2.4
Directional coupler measurements not for compliance assessment
For diagnostic purposes only, it may be helpful to take a directional coupler sample from the output of the
transmitter or a sample from the common point bridge. If the sample is taken from the bridge, set the
reactance (X) and resistance (R) dials detuned as far as possible so as to not place a null in or tip the
response of the spectrum of interest as observed at the common point's bridge detector's output.
The station should first be set up and confirmed to be mask-compliant in the field. Then, a transmitter
output or common point sample spectrum plot can be taken as a baseline reference value and saved for
future reference. For diagnostic purposes, a quick check of the transmission system performance can be
made by comparing a new transmission line or common point sample with the original on file. These
samples are not intended to verify mask compliance, but to show whether there has been a change in the
characteristics of the signal at this point in the transmission system since it was first set up.
5.2.3
Analog signal and digital sidebands
Establishing the carrier reference level is a vital step in measuring mask compliance. FCC rules
contained in 47 CFR §73.44 require the use of the non-directional power received at a measurement site
to be the reference carrier level for analog AM mask measurements. This has always been problematic
because it imposes a carrier-to-interference ratio penalty on the main lobe of a directional array that is
equal to the gain of the array in the main lobe.
For the purposes of hybrid AM IBOC mask compliance, it is the actual directional carrier level received at
the measurement point in the main lobe that establishes the carrier reference. This is because the welltransmitted hybrid AM IBOC signal has digital carriers that are quite close to the mask levels without
exceeding them. Using a non-directional carrier reference on the main lobe of the pattern would cause
the digital subcarriers to fall below their specified target levels for optimum performance.
Page 29
NRSC-G201-B
Spectral regrowth can be caused by non-linearities in the amplification system, including poorly aligned
magnitude and phase inputs, bias errors, and reflections caused by mismatches between the transmitter
and the antenna system (see Figure 11). Spectral regrowth occurs out of the main passband of the
antenna system and is likely to be attenuated by the antenna system to some degree. It is therefore often
beneficial to also examine the transmitter output as a diagnostic tool for evaluating how the transmitter is
behaving as part of the larger complex system.
Figure 11. AM IBOC spectral regrowth can be caused by non-linearities in the amplification
system, bias errors, and reflections caused by mismatches between the transmitter and antenna
systems. (Courtesy of Broadcast Signal Lab)
Spectral regrowth, however, can appear to be amplified by the antenna system on some radials from
directional arrays. This occurs when there are variations in the pattern bandwidth on certain radials
where the main carrier is attenuated more than the sidebands. Compliance, however, is not dependent
on the off-axis performance of the hybrid AM signal. Field evaluation for compliance only occurs in the
main lobe of the directional array. The directional station may need to address an off-axis emission in
response to specifically identified interference issues, on a case-by-case basis.
See Section 6.3 on measuring for further information.
5.2.4
Spurious emissions
AM stations are capable of receiving the emissions of other AM stations and then transmitting the
resulting intermodulation products. These spurious emissions can be evaluated anywhere around the
non-directional station and on the main lobe of the directional array, provided the receiving antenna has a
known frequency response at the frequency of the spurious emission.
Spurious emissions are best observed using “peak-hold” measurements such as those specified in 47
CFR §73.44.
Page 30
NRSC-G201-B
5.2.5
AM IBOC mask compliance measurement procedures
To harmonize the use of the 47 CFR §73.44 analog peak mask (“legacy mask”) and the NRSC-5 hybrid
IBOC PSD masks (shown in Figure 8 and Figure 9), the following protocol is recommended. If this
protocol conflicts with present or future FCC rules or enforcement policy, the broadcast engineer should
conform to the legal obligations of the station. Note that when performing AM IBOC mask compliance
measurements, normal program modulation should be used.
NRSC-5-C reference document 1082s, Section 4.4, calls for a three-step procedure for verifying that an
hybrid AM IBOC transmission is in compliance with the relevant RF masks. These measurements should
be repeated (as appropriate) for each mode of station operation.
The first step is to verify that the analog-only portion of the transmission is in compliance with 47 CFR
§73.44 (the legacy mask). This measurement is done using “peak-hold” measurement techniques.
The second step (also specified in Section 4.4 of NRSC-5-C reference document 1082s) calls for a hybrid
IBOC mask measurement to be performed on the analog-only signal, with the mask extended at -65
dBc/300 Hz from 5 to 20 kHz offset for ±5 kHz analog operation (and to -65 dBc/300 Hz from 8 to 20 kHz
offset for ±8 kHz operation). This measurement is done using “averaging” measurement techniques.
Note that if peak-hold measurements were used in this step, the relative amplitude of the digital
sidebands with respect to the mask (and also to the analog host) would be different than for averaging
measurements.
The third step involves verifying that the full AM hybrid signal (with both digital and analog components) is
compliant with the relevant hybrid AM IBOC mask (either Figure 8 or Figure 9). This third step is called
out in Section 4.5 of NRSC-5-C reference document 1082s.
To perform these measurements, the following sequence is recommended:
1) Legacy analog mask measurement – perform a peak-hold measurement of the analog-only
signal. The hybrid IBOC transmitter should be operating with the digital sidebands temporarily
disabled. This measurement follows the procedure specified under 47 CFR §73.44. Employ the
§73.44 mask, activate peak hold on the measuring instrument at 300 Hz resolution bandwidth,
collect data under normal program modulation for 10 minutes.
Performing this measurement ensures that any impulsive noise generated by the analog
transmission is captured and compared to against the FCC legacy analog mask.
2) Analog-sideband-to-digital overlap measurement – with the analog-only signal still
transmitting, perform a power averaging measurement with the hybrid IBOC mask measurement
method. The hybrid IBOC transmitter should be operating with the digital sidebands temporarily
disabled. Verify that the analog-only signal of a ±5 kHz-limited analog transmission does not
exceed -65 dBc/300 Hz PSD (power averaging) between the 5 kHz offset and the 20 kHz offset
above and below the carrier frequency. If the analog signal is set to the ±8 kHz bandwidth, verify
that the analog-only signal does not exceed -65 dBc/300 Hz PSD (power averaging) between the
8 kHz offset and the 20 kHz offset above and below the carrier frequency.
Performing this measurement ensures that the analog transmission meets the NRSC-5-C
specification for self-interference from the analog to the digital components of the hybrid IBOC
signal within the ±20 kHz bandwidth of the station.
3) Full Hybrid IBOC Measurement – Restore the full hybrid IBOC signal (analog and digital
components) and perform a hybrid IBOC mask measurement. This is the measurement that is
specified in Reference [1], Section 4.5 and explained further above. Use the 300 Hz resolution
bandwidth and power averaging function. Follow the procedure described above for the 100
sweep/30 second time period measurement.
Page 31
NRSC-G201-B
This measurement ensures that the hybrid IBOC signal and its spectral regrowth components
remain within the NRSC-5-C mask.
5.2.5.1
Additional legacy mask on full hybrid signal measurement
The NRSC-5-C standard specifies the three measurements above. This additional measurement is
helpful as a diagnostic tool, for instance, if impulsive noise is observed while the hybrid AM IBOC signal is
transmitting.
To do this additional measurement, perform a §73.44 peak-hold mask measurement on the full hybrid AM
IBOC signal (with both digital and analog components present). This measurement is useful because of
the complex interactions which occur between the analog and digital components of the hybrid AM IBOC
signal – see Annex 6 for additional information.
5.2.5.2
Harmonics
When measuring harmonics with a spectrum analyzer (instead of a narrowly tuned field intensity meter), it
is helpful to insert an AM band trap filter whose insertion loss and response at the harmonic frequencies
is known. This reduces the power of the incoming fundamental frequency, reducing the possibility that
the analyzer’s internal harmonic distortion will be mistaken for a transmitted harmonic product. A
calibrated antenna (or calibrated sampling port on a transmission line, if applicable) is necessary for
accurate measurements of harmonics. The NRSC-5 spectral mask applies to harmonic measurements of
hybrid IBOC AM signals.
5.2.5.3
Testing for instrument-induced harmonic and spurious content
Another valuable tip for studying harmonic and intermodulation products on an analyzer involves testing
an apparent spurious emission for whether it is emitted by the system under test or internally generated
by the measurement instrument. If the instrument is being forced into a non-linear range (due to analyzer
input overload) that causes internal spurious products, the addition of attenuation to the input of the
instrument will change the spurious component by more than the value of the attenuator.
For example, with the insertion of a 10 dB attenuator, all external signals entering the instrument will be
attenuated by 10 dB. However, any internally generated spurious signals will diminish by more than 10
dB, because the internal process is not linear. This is a helpful test when confronted with a spur and the
performance of the instrument is uncertain.
5.3
Troubleshooting (various out-of-spec conditions and possible remedies)
Before the hybrid AM IBOC transmission is evaluated for RF mask compliance, a number of initial design
and installation steps are required to optimize system performance, including:
●
The antenna system should be tuned, and if necessary, redesigned to be tuned, to provide the
impedance bandwidth and Hermitian symmetry desired of a well-running AM IBOC system. See
Annex 5;
●
The transmitter impedance matching filter (output network) rotates the passband on the Smith
Chart and must be accounted for when designing the antenna system characteristics and taking
impedance measurements;
Page 32
NRSC-G201-B
●
Transmitters having phase and magnitude inputs should be adjusted per manufacturer
specifications to obtain the ideal phase/magnitude input alignment;
●
The output of the hybrid AM IBOC transmitter should be as clean as possible to ensure that the
signal gets off to the best start possible.
5.3.1
Digital Signal Quality Testing
Digital waveforms can be analyzed for their deviation from the corresponding ideal waveform. This is a
powerful technique to characterize the overall “health” of a transmitted signal and help troubleshoot
problems with transmitted IBOC signals. Distortions of the waveform increase the chance of receiving
errors in the symbols used to represent the digital information. These distortions can be measured and
the results can be compared to recommended specifications.
Timing, amplitude, and phase distortions encountered in the transmission plant, the RF channel, and the
receiver all contribute to errors in reception of the digital radio signal. Engineers try to minimize the
distortion in the transmitted signal in order to reserve an error margin for the distortions contributed by the
RF channel and the receiver. In other words, by transmitting as distortion-free a signal as is reasonably
possible, the broadcaster maximizes the error correction “headroom” available at the receiver for the
impairments caused by the propagation and reception of the digital signal. By making digital signal
quality measurements at the transmission plant, the amount of headroom available can be characterized.
Further, these measured characteristics can help identify problems in the transmission system in cases
where headroom is found to be inadequate.
5.3.1.1
Modulation Error Ratio (MER)
One method for performing digital signal quality testing on an AM IBOC signal is to measure the
9
Modulation Error Ratio (MER). MER is a measurement of the signal-to-noise ratio (SNR) of the binary
information carried on the digital waveform. MER measurements tend to give the broadcast engineer a
more useful, “grayscale,” diagnostic view of system problems than the more “black and white” view
obtained from bit error ratio measurements (discussed in 4.4.3.2). The specific manner in which MER is
measured depends upon which subcarriers—data subcarriers or reference subcarriers—are used for the
10
measurement.
As with SNR, a higher MER indicates a better quality signal. As the MER decreases, the likelihood of
errors in reception increases.
5.3.1.2
Bit Error Ratio
In the digital domain, it is also customary to demodulate the digital waveform and evaluate the accuracy
of the resulting bit stream compared to the original. This can be reported as the bit error ratio (BER),
reporting the ratio of the number of bit errors to the total number of bits being evaluated. The “raw,”
uncorrected BER measures the demodulated data stream before error correction is applied in the
receiver. There is also a post error correction BER that applies the digital system’s built-in error
correction features and identifies the bit errors that remain after error correction.
The post-error correction BER in the receiver tends to exhibit a cliff effect, in that the output remains
relatively error-free over a wide range of input SNR, but then at some point the SNR reaches a
“threshold” after which a relatively small change in SNR overwhelms the system with errors and the
9
Currently there is no specification for MER measurement of an AM IBOC signal, but such a specification may be
developed in the future. A specification for FM IBOC MER measurement may be found in [20].
10
See [1] for additional information on the structure of an AM IBOC signal.
Page 33
NRSC-G201-B
output error ratio then rises steeply, with an abrupt system failure ensuing. For this reason, a post-error
correction BER will tend to mask subtle impairments in the signal under test, while the raw BER provides
more information about the number of errors occurring before the system fails.
5.3.1.3
MER vs. BER
There is somewhat of a paradox in the reception and characterization of received digital signals. The
measurement device must acquire the signal in order to measure it, but the act of acquiring relies on
aligning the measurement device to the signal under test. In addition to the error correction (which affects
the post-error correction BER), the digital signal receiver first must perform several steps to “lock on” to
the signal. These steps include frequency adjustment to center the receiver on the exact center
frequency of the waveform, timing adjustment to lock onto the symbols and identify the proper starting
point of each symbol, and phase equalization to adjust for the variation of the phase of the digital
waveform across the occupied channel. The pre- and post-error correction BER measurements rely on
the measurement device to perform these adjustments to acquire the signal.
Different receiver designs may behave differently under certain kinds of signal impairments and it is
therefore important for BER measurements to be performed with a uniform reference receiver design. By
contrast, MER measurements made at the transmitter may use simpler and more consistent algorithms
since there will be no interfering signals, Doppler shift, propagation effects, multipath, etc. If the MER
measurement is made at the IBOC exciter, the frequency and timing information is already, directly,
available along with both the original data and the demodulated data sample from the output of the
transmission system. MER can measure, among other aspects, the unequalized phase performance of
the signal across the channel bandwidth, whereas the BER relies on the receiver doing some work to
correct phase to the best of its ability. By taking an MER measurement at the antenna transmission
system output to the antenna, the effects of the propagation channel and the effects of the receiver
equalization are removed from the measurement, revealing only the effects of the transmission system on
the waveform.
Page 34
NRSC-G201-B
6
Measurement Methodology
6.1
6.1.1
Spectrum analyzers
Input levels
Total broadband RF power into the instrument’s mixer, detector, or converter should be well below its 1
dB compression level to allow for peaks on desired and undesired signals to be handled without distortion
and without resulting in internal generation of false signals. When evaluating the actual input level to the
instrument, not only must the total RF power of all signals present on the sample line be considered, but
also the internal attenuator setting should be accounted for.
Spectrum analyzers have internal input attenuators that can be manually or automatically engaged. The
instrument operator should keep in mind that analyzers correct displayed readings to compensate for the
value of the internal attenuation. This means that the operator must calculate the total power reaching
the input mixer to determine how much mixer headroom there is.
If, for example, the total power delivered to the instrument’s input connector were 0 dBm, and the internal
attenuator is set to 10 dB, then the input power to the first stage is -10 dBm. This is true even though the
analyzer reports the level as 0 dBm. In this example, after determining the total input power to the first
active stage of the analyzer is -10 dBm, the headroom can be calculated. If the 1 dB compression level
of the analyzer were +7 dBm, the addition of the 10 dB of internal attenuation gives the instrument a
comfortable 17 dB headroom between the total input power (average) to the first stage and the 1 dB
compression level.
The operator should keep in mind that IBOC digital signals, in the absence of the analog signal, may have
peak to average ratios in the vicinity of 4-6 dB, which should be considered in ensuring that there is
sufficient headroom for an accurate measurement. Likewise, undesired signals also arriving at the input
to the instrument may contribute average and peak power levels that must be accounted for in setting the
input level.
6.1.2
Dynamic range
The noise floor of instrumentation being used for mask compliance measurement should be at least 10
dB below the mask minimum for best results. For example, the most restrictive level of the hybrid AM
IBOC mask is -85 dBc. The instrument noise floor should be therefore -95 dBc or lower. This prevents
instrument noise from summing with the measured value to create a false appearance of non-compliance.
Spectrum analyzer performance is characterized by various measures. The first measures to examine
are the displayed average noise level (DANL) and the 1 dB compression level. Be aware that the DANL
varies with frequency, requiring careful attention to the DANL frequency and its resolution bandwidth
settings. When scaled to the desired resolution bandwidth for the measurement, the factory DANL should
be low enough to offer the desired dynamic range. Note that contemporary spectrum analyzers typically
have published specifications for noise levels and other characteristics beginning at 10 MHz, and above.
Instrument noise levels may be higher at frequencies lower than 10 MHz, requiring verification by
demonstration that the instrument’s performance in the AM band meets the dynamic range requirements.
To determine the noise floor requirements for a mask compliance measurement device, use Equation 3:
Equation 3 (decibel addition & subtraction)
Page 35
NRSC-G201-B
(1 dB compression level)
– (margin to handle peaks without compression)
– (difference between total input power and input power of signal under test)
– (greatest mask clearance from signal under test reference level to lowest value of mask)
– (noise floor margin to resolve space below mask)
= (required noise floor)
Careful reading of the instrument specifications, while accounting for resolution bandwidths used in the
specifications, may reveal whether the instrument has the minimum necessary dynamic range. As
Equation 3 shows, if the input level of the signal under test is below the 1 dB compression level by a very
large margin, then the instrument noise floor must be correspondingly lower. This may occur if the
instrument is receiving other strong signals in addition to the signal under test or if the signal under test is
weak and cannot be brought up to the desired input level.
Considering that, for AM measurements, it is sometimes difficult to get a strong signal at a field test point,
the actual input level of the signal to the analyzer may need to be substantially lower in level than is
optimal. Similarly, for FM measurements there may be multiple strong signals on, for instance, a master
transmission line, so the input level of the signal under test may need to be lower than in the example
above, to account for the total power of all signals and the peak power of those signals appearing at the
instrument’s input. In all these cases, additional instrument dynamic range or low noise input
amplification will be necessary to accommodate weaker desired signal levels.
To illustrate Equation 3, assume a particular instrument has a +7 dBm 1 dB compression level. We want
to allow plenty of room for the peaks of multiple incoming signals to sum without affecting the
measurement, so in this example a headroom margin of 15 dB is selected to be sure that the
simultaneous occurrence of peaks of several incoming signals have some headroom. Assume that the
desired signal is one of 6 equal power signals on a master antenna system, which means the average
power of the desired signal is 8 dB down from the total input power (= 10* log [1/6]). The hybrid FM IBOC
mask’s lowest limit line is -80 dBc. As noted above, this Guideline recommends at least 10 dB between
the mask minimum and the instrument noise floor. Using Equation 3:
+7 dBm
(1 dB compression level)
– 15 dB
(margin to handle peaks without compression, selected on case-by-case basis)
– 8 dB
(difference between total input power and input power of signal under test)
– 80 dB
(greatest mask clearance from signal under test reference level to lowest value of mask)
– 10 dB-kHz (noise floor margin at desired resolution bandwidth to resolve space below mask)
= -106 dBm/kHz
(required noise floor)
So for this example, -106 dBm/kHz will be the maximum acceptable displayed average noise level. Note
that the required instrument noise floor calculated in the example may not be directly comparable to the
manufacturer’s specification. For instance, a specification might indicate that an instrument has a -130
dBm/Hz noise floor at the desired frequency. Increasing the bandwidth from 1 Hz PSD to 1 kHz
increases the noise by a factor of 1000, or 30 dB. Consequently, the instrument has a displayed average
noise level of -130 dBm/kHz + 30 dB = -100 dBm/kHz.
Other instrument performance measures include intermodulation specifications in several forms. It is not
the role of this document to explain these specifications in detail, only to make the reader aware of them.
An instrument may have a satisfactory noise floor, but the intermodulation specifications must be good
enough that its ability to identify real spurious emissions of a hybrid IBOC signal is not compromised by
internally generated products. There is no better way to evaluate an instrument than to try it out on
signals whose performance is already known.
Page 36
NRSC-G201-B
6.1.3
Resolution bandwidth (RBW) and noise bandwidth
Resolution bandwidth (RBW) is one of the most important parameters to be selected in setting up the
spectrum analyzer. RBW can be thought of as the amount of bandwidth that is analyzed for each
measurement point on a spectrum analyzer trace. RBW is typically established using selectable
bandwidth filters within the analyzer. Most analyzers can select the RBW automatically, usually as a
function of the frequency span being analyzed, and in addition provide for manual selection of RBW.
When making IBOC mask compliance measurements, it is customary to use a 1 kHz RBW for FM IBOC
and a 300 Hz RBW for AM IBOC since these are the bandwidths used in expressing the masks (Figure 1,
Figure 8, and Figure 9). However, there may be a reason for employing a different RBW, as explained in
more detail in Annex 1.
RBW filters have associated with them a particular noise bandwidth. Noise bandwidth refers to the
effective bandwidth of a filter (or other device) when measuring noise or noise-like signals. Noise
bandwidth affects measurements of the OFDM digital sidebands because OFDM signals are noise-like.
For example, two filters may have the same 3 dB bandwidth but have different width on the skirts. The
filter with wider skirts will collect more “noise” power from the digital signals being measured than will the
filter with narrower skirts, and hence will have a larger noise bandwidth.
The power of a digital signal reported by the spectrum analyzer for a given RBW is actually overstated by
some error value. The amount of overstatement is a function of the noise bandwidth of the type of filter
used in the analyzer. To obtain a correct power reading from a spectrum analyzer display, the effect of
the noise bandwidth on the measurement of a digital signal should be subtracted:
●
●
●
4-pole synch tuned analog filters: subtract 0.5 dB from readings
Gaussian analog filters: subtract 0.24 dB from readings
Gaussian filters implemented digitally: inconsequential (0.01 to 0.02 dB)
Some (newer) instruments can account for the RBW filter noise bandwidth when measuring digital
signals. The operator should be as familiar as possible with the features of the instrument.
6.1.4
Detectors
The noise-like nature of modulated OFDM waveforms affects certain analyzers. Older units that use a
“sample” detector and simply average the trace data on the log display implicitly understate the power by
approximately 2.5 dB. Newer units may appear to average the trace but could be averaging the power
data from which the trace is derived, without the 2.5 dB error. Newer units may also have “average
power” detectors, (also called “RMS” detectors) that accurately measure digital signal power. Some
instruments have an “average” detector that is not an average power detector, but a max/min averaging
method applied to the maximum and minimum value in each measurement bin. Consult the instrument
manual or the manufacturer to learn about the detectors available.
Alternatively, rather than adjusting the reading for this 2.5 dB error, the mask can be adjusted when
creating a reference level (or limit line) on an instrument to take this into account. The correction may be
subtracted from the published mask or added to the reading. For instance, an old analyzer with 4-pole
filter, sample detector, log trace averaging would have 2.5 – 0.5 = 2.0 dB added to the reading (by
lowering the reference level 2.0 dB below the carrier level), or the same amount would be subtracted from
the mask overlaid on the instrument as a limit line.
It is recommended that a spectrum analyzer with an average power or RMS detector be obtained to
eliminate the need for the 2.5 dB correction. For improved accuracy, the RBW filter noise bandwidth still
should be accommodated, unless the unit employs digital filtering.
Page 37
NRSC-G201-B
Traditional analog measurements under the FCC §73.44 legacy mask are performed with the peak
detector, 300 Hz RBW, and an accumulation of 10 minutes of sweeping on a display set to maximum
hold.
6.1.5
Limit lines
Limit lines are a convenient feature for evaluating the compliance of a measured signal with its mask.
Contemporary spectrum analyzers have the ability to draw limit lines that represent spectral masks (see
limit lines, in green, shown in Figure 12). Consult with the instrument’s manual for instructions on how to
program, store, recall, and set a reference level for limit lines. Some instruments will also indicate on the
display whether the measured spectrum passes or fails the limit line test.
Figure 12. Limit lines are a powerful tool and can greatly simplify mask compliance
measurements. (Courtesy Burt Weiner Associates and Broadcast Signal Lab)
6.2
6.2.1
FM setup – spectrum analyzers
Setting frequency span
The frequency span for Hybrid FM IBOC measurements is most conveniently set to about 500-600 kHz to
provide sufficient detail of the hybrid IBOC signal while maintaining a full view of the hybrid IBOC signal.
The center frequency may then be offset to each side of the display to look at upper and lower sideband
spectral regrowth and spurs with similar detail. Alternatively, to view the hybrid FM IBOC signal and at
least the first digital spectral regrowth energy at ±492 kHz offsets, use 1.2-2 MHz span with the carrier at
center frequency.
Page 38
NRSC-G201-B
6.2.2
Resolution bandwidth (RBW)
The default RBW for FM IBOC measurements is 1 kHz. The characteristics of the RBW filter may slightly
affect the results (see Section 6.1.3 above). Assuming the spectrum analyzer employs the traditional
four-pole synchronous filters or filters with narrower skirts, the hybrid FM IBOC mask provides plenty of
room between the response of common 1 kHz RBW filter skirts and the mask slope at 200-250 kHz.
However, if a station routinely has difficulty making the mask on this slope with a traditional analyzer with
4-pole analog filters, the station may do one of two things to improve its chances:
●
First, stations may employ a narrower RBW, such as 100 Hz, 10 Hz, or even 1 Hz. The resulting
measurements, at 100 HZ RBW, will be 100/1000 lower in level (i.e., 10 dB), so either the mask
must be lowered 10 dB (as a limit line on the instrument) or the measurement must be
compensated by increasing the result by 10 dB. Either way, the use of the narrower resolution
bandwidth may improve the response on the slopes and determine that the mask has indeed been
satisfied;
●
The second alternative to addressing a mask-slope compliance problem is to employ a spectrum
analyzer with narrower shape-factor filters. Gaussian filters (digital or analog) operated at the 1
kHz RBW may be sufficient to confirm the signal is within the mask. The goal of making such
changes to the measurements is to minimize the error contributed by the instrument in order to
verify compliance of the signal with the mask.
Alternatively, there may not be a mask-slope issue. If so, it may produce results more quickly to use a
wider RBW as long as it confirms compliance and does not create the appearance of failure. With a 4pole analog filter, the widest practicable RBW for FM mask measurements is 3 kHz. Any wider bandwidth
competes with the slope of the mask. With a digital Gaussian filter, a bandwidth slightly wider than 6 kHz
may be employed. The mask must be adjusted to compensate for the change in bandwidth and any
change in the filter noise bandwidth should be accommodated. For example, using a Gaussian 6 kHz
RBW filter, the mask should be adjusted up by 10 x Log(6 kHz/1 kHz) = 7.78 dB
In Annex 1, an explanation of how the power spectral density measurements of IBOC signals can be
normalized to an equivalent power spectral density per Hz is given. Even if the spectrum analyzer or
other instrument cannot measure to 1 Hz resolution, such normalization permits use of virtually any
RBW, as long as the appropriate adjustments to the mask are made.
6.2.3
Video bandwidth
The video filter smooths out the sweep as it displays, potentially robbing the operator of a view of the fine
structure of the occupied spectrum. It is suggested that to observe gain flatness and repeatable spiky
behavior, the video filter is best turned off or set to a value that is at least ten times the resolution
bandwidth (e.g., 10 kHz for a 1 kHz RBW setting).
6.2.4
Detector
Use “average power” or “RMS” detector if available. If not, use a “sample” detector with trace averaging.
Do not use an “average” or “mean” detector that averages the max and min value in each bin. Peak
detection should not be used for IBOC mask compliance measurements.
6.2.5
Reference level
Page 39
NRSC-G201-B
Ideally, an unmodulated analog FM carrier is used to establish the reference level for determining mask
compliance. Set the peak of the unmodulated carrier to the top of the spectrum analyzer display; having
done so, the top of the display becomes the 0 dBc reference level. To conveniently identify the peak and
set it to the reference level, use the marker function on the spectrum analyzer, if available; use the “find
peak” function, then select “set marker to reference level”.
If modulation cannot be removed, use a wide RBW (300 kHz for FM IBOC is a good choice) and
determine the power level at the center frequency when modulation is present (Figure 13). Conducting
this measurement over numerous sweeps is not necessary, but will help remove what little noise there is
in the reference level. Then, set the power level measured at the center frequency (using the wide RBW)
to the reference level at top of display; the top of the display now corresponds to 0 dBc. There is a minor
error due to the presence of the IBOC signals within the passband of the 300 kHz RBW at center
frequency, but it is inconsequential (0.04 dB at most).
Figure 13. When a modulated signal is used to set the 0 dBc reference level, use a wide resolution
bandwidth setting (300 kHz is used here), then use the power level at the center frequency to
establish the reference level. (Courtesy of Broadcast Signal Lab)
Note that three different sweeps are shown in Figure 13:
●
●
●
6.2.6
the unmodulated carrier (lower/brown sweep);
the modulated carrier plus IBOC with a 1 kHz RBW (middle/blue sweep);
the modulated carrier plus IBOC with a 300 kHz RBW (upper/green sweep).
Sweep rate and number of sweeps
Page 40
NRSC-G201-B
The sweep rate will be optimized if the instrument is left on “Auto” sweep as the RBW is selected. When
setting manually, the rule of thumb is to take one-half of the square of the RBW as the target sweep rate.
At 1 kHz RBW, the optimum sweep rate is approximately 500 kHz per second (instruments may calculate
faster rates based on their preset parameters; these may be relied upon). Sweeping too quickly will
cause the dynamics of the signal on each sweep to be missed. Sweeping too slowly has no
measurement consequence. With old instruments and sample detection, more sweeps yield a better
average, and sweeping more slowly than necessary merely wastes time. However, newer instruments
with average power or RMS detectors are efficient at any rate slower than the maximum Auto rate. They
may be swept very slowly, taking a long-term trace, or an average of a few traces; or they may be set to
sweep more quickly to average many quick traces.
The goal is to take a measurement of the signal over a period that is long enough to minimize the
variance between successive measurements. The benefit of averaging multiple sweeps is that the
operator can watch the average accumulate and then determine visually when the result has stabilized.
The NRSC-5 transmission system specification calls for 100 sweeps. The sweep rate formula (or Auto
Sweep mode) determines the minimum time required to obtain 100 sweeps. Set up in this fashion,
measurements will certainly have a low variance. This is an effective way to document results. For faster
analysis and examination work, the operator can limit the number of sweeps to that which is necessary to
observe and make adjustments. Also, for quicker results, a narrower sweep bandwidth (“span” on a
spectrum analyzer) reduces the time necessary to complete a sweep.
The NRSC-5 transmission system specifications also call for a minimum of 30 seconds of data
accumulation on the spectrum analyzer. By using the highest reliable sweep rate and the narrowest span
necessary for a period of time that includes at least 100 sweeps, the elapsed time may be more or less
than 30 seconds, depending on the span utilized. To comply literally with the specification, there should
be at least 100 sweeps and a collection time of at least 30 seconds.
As a practical matter, experience has shown that 100 sweeps taken at the maximum applicable sweep
rate is sufficient for obtaining a low-variance measurement, regardless of the time required to complete
the sweeps. Thus, for analyzers set to sweep at the auto rate – at most span settings – the number of
sweeps tends to be the controlling factor above the exact duration of the measurement.
Alternatively, for instruments that can perform accurate averaging without requiring multiple sweeps (e.g.,
average power/RMS detection and FFT-based analyzers), it is acceptable to determine by
experimentation how much time is required to perform a measurement that minimizes the variance
between measurements. With these instruments, the number of sweeps may be one or very few, in
which case the duration of the measurement is the controlling factor over the number of sweeps.
6.2.7
Interpreting results
There will always be some degree of variation between two measurements, whether due to different
instruments, different operators, or measurements taken at a different time. The greatest source of
spectrum analyzer uncertainty in IBOC RF mask measurements is in the detection and averaging
functions. These are covered in detail in this Guideline.
The amplitude uncertainties of spectrum analyzers are of less concern. Specifications for spectrum
analyzers provide several figures to describe various amplitude uncertainties present in the apparatus,
including among others, frequency response, band switching, IF gain, RBW, and attenuator switching
uncertainties. Fortunately, most of these uncertainties are minimized or eliminated because IBOC mask
measurements are relative measurements. The reference analog carrier power is very close in frequency
to the IBOC spectrum under test and there is no need to switch the instrument input characteristics once
the input level and bandwidth has been set.
One exception is when the analog FM signal power is obtained on a modulated signal with a 300 kHz or
wider RBW, followed by RBW switching to 1 kHz for the mask measurement. In this case, the RBW
Page 41
NRSC-G201-B
switching uncertainty applies. Typically, it is reasonable to expect that the amplitude uncertainty of a wellperforming analyzer performing IBOC RF mask measurements will be well within ±2 dB, with the results
being very likely to be within ±1 dB.
Another cause of uncertainty in measurements against the FM IBOC RF mask is the quality of the source.
The response of the directional coupler used to obtain the FM IBOC signal sample may affect the results,
particularly of out-of-band spurious emissions. In the case of separate-line systems, the estimates of
gains and losses on the digital and the analog transmission paths are likely to have discrepancies on the
order of 1-2 dB that confound accurate evaluation of the analog-to-digital power ratio. Finally, the
presence of unwanted energy from other sources can corrupt the results. Such energy could be
introduced through reverse path crosstalk in the coupler, test cable pickup of radiated or conducted
energy, and potential instrument-generated intermodulation and harmonic products.
A well-conducted and reported FM IBOC mask compliance measurement will include documentation of
the results, the methodology, and the uncertainties. The results should include both numeric and graphic
presentations of measurement data. The methodology should include procedure and equipment
descriptions as well as set-up diagrams. The uncertainties should include discussion of the potential
sources of measurement error and any special challenges faced in taking the measurement.
6.3
6.3.1
AM setup – spectrum analyzers
Setting frequency span
The frequency span for hybrid AM IBOC measurements is most conveniently set to about 40-50 kHz to
provide sufficient detail of the hybrid IBOC signal. The center frequency may then be offset to look at
upper and lower sideband spectral regrowth and spurs with similar detail. Alternatively, to view the hybrid
AM IBOC signal and at least the first spectral regrowth energy at ±25 kHz offsets, use 60-100 kHz span
with the carrier at center frequency.
6.3.2
Resolution bandwidth (RBW)
The default RBW for AM IBOC measurements is 300 Hz. The characteristics of the filter may affect the
results (see Section 6.1.3 above. Typically, spectrum analyzers employ four-pole synchronous filters or
filters with tighter skirts. The hybrid AM IBOC mask is quite close to the slope of the 4-pole
synchronously tuned RBW filter in the 15.0 to 15.2 kHz region. It may seem contradictory that this slope
is 200 Hz wide and is being measured by a 300 Hz RBW filter. However, because this is a power
spectral density measurement, and there are likely to be several data points (measurement bins) from
15.0 to 15.2 kHz, the analyzer is integrating the power in a 300 Hz bandwidth at each data point. The
mask slope was selected to insure that under ideal conditions the hybrid IBOC signal would register as
compliant with the mask with a 300 Hz RBW. However, if a station routinely has difficulty making the
mask on this slope with a traditional analyzer with 4-pole analog filters, the station may do one of two
things to improve its chances.
●
First, stations may employ a narrower RBW, such as 100 Hz, 10 Hz or even 1 Hz. The resulting
measurements, at 100 Hz RBW, will be 100/300 lower in level (4.8 dB), so either the mask must be
lowered 4.8 dB (as a limit line on the instrument) or the measurement must be compensated by
increasing the result by 4.8 dB. Either way, the use of the narrower resolution bandwidth may
improve the response on the slopes and determine that the mask has indeed been satisfied;
●
The second alternative to addressing a mask-slope compliance problem is to employ a spectrum
analyzer with narrower shape-factor filters. Gaussian filters (digital or analog) operated at the 300
Hz RBW may be sufficient to confirm the signal is within the mask. The goal of making such
changes to the measurements is to minimize the error contributed by the instrument in order to
verify compliance of the signal with the mask.
Page 42
NRSC-G201-B
In Annex 1, an explanation of how the power spectral density measurements of IBOC signals can be
normalized to an equivalent power spectral density per Hz is given. Even if the spectrum analyzer or
other instrument cannot measure to 1 Hz resolution, such normalization permits use of virtually any RBW,
as long as the appropriate adjustments to the mask are made.
6.3.3
Video bandwidth
The video filter smooths out the sweep as it displays, potentially robbing the operator of a view of the fine
structure of the occupied spectrum. It is suggested that to observe gain flatness and “spiky” behavior, the
video filter is best turned off or set to a value that is at least ten times the resolution bandwidth (e.g., 3
KHz for 300 Hz RBW setting).
6.3.4
Detector
Use “average power” or “RMS” detector if available. If not, use a “sample” detector with trace averaging.
Do not use an “average” or “mean” detector that averages the max and min value in each bin. Peak
detection should not be used for IBOC mask compliance measurements.
For traditional NRSC-2 measurements of analog-only AM signals to determine compliance with 47 CFR
§73.44, use the peak detector and the peak hold functions of the analyzer.
6.3.5
Reference level
Ideally, an unmodulated analog AM carrier is used to establish the reference level for determining mask
compliance. Set the peak of the unmodulated carrier to the top of the spectrum analyzer display; having
done so, the top of the display becomes the 0 dBc reference. To conveniently identify the peak and set it
to the reference level, use the marker function on the spectrum analyzer, if available; use the “find peak”
function, then select “set marker to reference level”.
If modulation cannot be removed, use a narrow RBW (300 Hz or less for AM IBOC is a good choice) and
perform an average power measurement of the modulated carrier over numerous sweeps (average until
subsequent averages do not change the level at center frequency significantly). When enough averages
have been taken, determine the power level at center frequency using the zero span mode of the
analyzer. Then, set the power level measured at the center frequency level to the reference level at top
of display; again, the top of the display now corresponds to 0 dBc (see Figure 14). There may be a minor
error due to the asymmetrical modulation (-99, +125%), but it is inconsequential, especially at a narrow
RBW.
Page 43
NRSC-G201-B
Figure 14. When a modulated AM signal is used to set the 0 dBc reference level, average the
analyzer display over many traces, then use the power level at the center frequency to establish
the reference level. (Courtesy of Broadcast Signal Lab)
6.3.6
Sweep rate and number of sweeps
The sweep rate will be optimized if the instrument is left on “Auto” sweep as the RBW is selected. When
setting manually, the rule of thumb is to take one-half of the square of the RBW as the target sweep rate.
At 300 Hz RBW, the optimum sweep rate is approximately 45 kHz per second (instruments may calculate
faster rates based on the set parameters; these may be relied upon). Sweeping too quickly will cause the
dynamics of the signal on each sweep to be missed. Sweeping too slowly has no measurement
consequence. With old instruments and sample detection, more sweeps yield a better average, and
sweeping slowly merely wastes time. However, newer instruments with average power or RMS detectors
are efficient at any rate slower than the maximum Auto rate. They may be swept very slowly, taking a
long-term trace or an average of a few traces; or they may be set to sweep more quickly to average many
quick traces.
The goal is to take a measurement of the signal over a period that is long enough to minimize the
variance between successive measurements. The benefit of averaging multiple sweeps is that the
operator can watch the average accumulate and then determine visually when the result has stabilized.
The NRSC-5 transmission system specification calls for 100 sweeps. The sweep rate formula (or Auto
Sweep mode) determines the minimum time required to obtain 100 sweeps. Set up in this fashion,
measurements will certainly have a low variance. This is an effective way to document results. For faster
analysis and examination work, the operator can limit the number of sweeps to that which is necessary to
observe and make adjustments. Also, for quicker results, a narrower sweep bandwidth (“span” on a
spectrum analyzer) reduces the time necessary to complete a sweep.
The NRSC-5 transmission system specifications also call for a minimum of 30 seconds of data
accumulation on the spectrum analyzer. By using the highest reliable sweep rate and the narrowest span
necessary for a period of time that includes at least 100 sweeps, the elapsed time may be more or less
Page 44
NRSC-G201-B
than 30 seconds, depending on the span utilized. To comply literally with the specification, there should
be at least 100 sweeps and a collection time of at least 30 seconds.
As a practical matter, experience has shown that 100 sweeps taken at the maximum applicable sweep
rate is sufficient for obtaining a low-variance measurement, regardless of the time required to complete
the sweeps. Thus, the number of sweeps implicitly takes precedence over the exact duration of the
measurement.
Alternatively, for instruments that can perform accurate averaging without requiring multiple sweeps (e.g.
average power/RMS detection and FFT-based analyzers), it is acceptable to determine by
experimentation how much time is required to perform a measurement that minimizes the variance
between measurements. With these instruments, the duration of the measurement implicitly takes
precedence over the number of sweeps.
6.3.7
Interpreting results
There will always be some degree of variation between two measurements, whether due to different
instruments, different operators, or measurements taken at a different time. The greatest source of
spectrum analyzer uncertainty in IBOC RF mask measurements is in the detection and averaging
functions. These are covered in detail in this Guideline.
The amplitude uncertainties of spectrum analyzers are of less concern. Specifications for spectrum
analyzers provide several figures to describe various amplitude uncertainties present in the apparatus,
including among others, frequency response, band switching, IF gain, RBW, and attenuator switching
uncertainties. Fortunately, most of these uncertainties are minimized or eliminated because IBOC mask
measurements are relative measurements. The reference analog carrier power is very close in frequency
to the IBOC spectrum under test and there is no need to switch the instrument input characteristics once
the input level and bandwidth has been set. Typically, it is reasonable to expect that the amplitude
uncertainty of a well-performing analyzer performing IBOC RF mask measurements will be well within ±2
dB, with the results being very likely to be within ±1 dB.
Another cause of uncertainty in measurements against the AM IBOC RF mask is the quality of the source.
In the field, the presence of energy generated by other sources may corrupt or mask the emissions of
interest to the tester. Careful selection of a measurement site (or sites, if necessary) will minimize the
potential interference. A directional antenna can be employed to isolate unwanted emissions in the
spectrum or at least identify the direction from which they are coming
In cases where transmitter output measurements are being used to verify RF mask compliance, the
response of the voltage and current vector sampling apparatus on the transmission line must be reliable
or it may affect the results. Directional couplers and RF sampling loops may unreliably sample the
spectrum in an AM transmission system and should only be used to benchmark performance for
diagnostic purposes and future reference rather than for mask compliance purposes.
A well-conducted and reported AM IBOC mask compliance measurement will include documentation of
the results, the methodology, and the uncertainties. The results should include both numeric and graphic
presentations of measurement data. The methodology should include procedure and equipment
descriptions as well as set-up diagrams. The uncertainties should include discussion of the potential
sources of measurement error and any special challenges faced in taking the measurement.
Page 45
NRSC-G201-B
7
Further developments
It is anticipated that instrument manufacturers may develop innovative methods for evaluating signals and
achieving compliance. They might not be in the form of traditional spectrum analyzers. Annex 1 of this
document provides some insight into the malleability of the measurement technique to accommodate new
methods while still maintaining compliance with the fundamental mask.
For Hybrid FM IBOC transmissions, higher digital power levels with respect to the analog host will likely
complicate some of the issues discussed in this Guideline. The combined hybrid IBOC signal on a
common transmission line will have a higher peak-to-average ratio than the original -20 dBc digital signal
power, affecting transmitter linearity as well as measurement instrument linearity. With higher digital
power levels comes greater potential for the digital output to crosstalk into the analog power amplifier (for
separate amplification systems) and for spectral regrowth caused by amplifier non-linearities (in common
amplification systems). The directivity of directional couplers employed to sample common line and
separate line hybrid IBOC transmissions may need to be increased to provide greater front-to-back
isolation for accurate measurements of PSD.
For hybrid AM IBOC transmissions manufacturers are already providing vector analysis of outgoing signal
power with some transmitter products. With increasing market penetration and increasing industry
experience with these devices, more precise and more useful measurements of AM signals will be
available to more hybrid AM IBOC facilities.
Page 46
NRSC-G201-B
ANNEX 1 – DISCUSSION OF A TRACTABLE APPROACH TO DEFINING AND MEASURING
IBOC SIGNALS AGAINST RF MASKS
Page 47
NRSC-G201-B
A Tractable Approach to Defining and Measuring IBOC
Signals against the RF Masks
By David Maxson, Broadcast Signal Lab, LLP, Medfield, MA, with significant conceptual and
editorial contributions by Joe Gorin, Agilent Technologies, Inc., Santa Rosa, CA. Further
contributions by the NRSC IBOC Standards Development Working Group. September 2008
Table of Contents
1.
INTRODUCTION ....................................................................................................................................... 49
1.1.
2.
GOALS OF THIS PAPER ............................................................................................................................... 50
TRACTABLE AND SUGGESTED OPERATIONAL SPECIFICATIONS FOR HYBRID IBOC
SPECTRAL OCCUPANCY MEASUREMENTS .................................................................................... 51
2.1.
2.2.
2.3.
2.4.
2.5.
2.6.
2.7.
3.
HYBRID FM DIGITAL SIGNAL FLATNESS AND OUT-OF BAND EMISSIONS: SPECIFICATION 1— TRACTABLE
DEFINITION ............................................................................................................................................... 51
OPERATIONAL VARIATIONS: BANDWIDTH ................................................................................................ 52
OPERATIONAL VARIATIONS: REFERENCE LEVEL ...................................................................................... 53
OPERATIONAL VARIATIONS: AVERAGING TIME ........................................................................................ 54
MEASURING HYBRID FM IBOC SPECTRAL OCCUPANCY AGAINST THE MASK: ........................................ 55
MEASURING HYBRID FM DIGITAL SIGNAL SUBCARRIER GROUP POWER (I.E. PRIMARY MAIN, PRIMARY
EXTENDED UPPER OR LOWER SIDEBANDS): ............................................................................................... 56
FURTHER DISCUSSION OF FM IBOC MEASUREMENTS WITH SPECTRUM ANALYZERS............................... 58
AM ................................................................................................................................................................ 60
3.1.
3.2.
3.3.
3.4.
3.5.
3.6.
3.2.
4.
HYBRID AM DIGITAL SIGNAL FLATNESS AND OUT-OF BAND EMISSIONS: SPECIFICATION 2 – TRACTABLE
DEFINITION ............................................................................................................................................... 60
OPERATIONAL VARIATIONS: BANDWIDTH ................................................................................................ 61
OPERATIONAL VARIATIONS: REFERENCE LEVEL ...................................................................................... 62
OPERATIONAL VARIATIONS: AVERAGING TIME ........................................................................................ 62
MEASURING HYBRID AM IBOC SPECTRAL OCCUPANCY AGAINST THE MASK: ........................................ 63
MEASURING HYBRID AM DIGITAL SIGNAL SUBCARRIER GROUP POWER (I.E. PRIMARY, SECONDARY, AND
TERTIARY UPPER OR LOWER SIDEBANDS): ................................................................................................. 65
FURTHER DISCUSSION OF AM IBOC MEASUREMENTS WITH SPECTRUM ANALYZERS .............................. 66
DISCUSSION OF SPECTRUM ANALYZER MEASUREMENTS ...................................................... 67
4.1.
4.2.
MAKING MEASUREMENTS TO THE STANDARDS WITH NEWER SPECTRUM ANALYZERS ............................ 67
MAKING MEASUREMENTS TO THE STANDARDS WITH OLDER SPECTRUM ANALYZERS: EXACT BUT
INCONVENIENT TECHNIQUES..................................................................................................................... 67
4.3.
MAKING MEASUREMENTS TO THE STANDARDS WITH OLDER SPECTRUM ANALYZERS: INEXACT BUT
CONVENIENT TECHNIQUE ......................................................................................................................... 68
4.3.1. How Close to Gaussian is the IBOC Digital Signal? .......................................................................... 68
4.4.
AVERAGING TECHNIQUES ......................................................................................................................... 69
4.4.1. Trace Averaging vs. Video Filtering .................................................................................................... 69
4.4.2. Long Sweep Versus Trace Averaging .................................................................................................. 69
Page 48
NRSC-G201-B
1. Introduction
The design of transmitters is constrained by the tension between cost and performance. With
respect to transmitting digital signals, the cost of ownership and operation increases with the
linearity required of the transmitter. This tension has a direct impact not only on the potential
quality of the transmitted signal, but also on the generation and radiation of unwanted energy in
the radio spectrum due to intermodulation. The In-Band/On-Channel (“IBOC”) radio format
known by the HD RadioTM brand name of iBiquity Digital Corporation 11 requires only such
transmitter linearity that is sufficient for transmitting a clean digital signal component in the
presence of analog signals on the same and nearby channels.
Experimentally, the peak-to-average power ratio (“PAPR”) of the digital IBOC FM signal at
baseband is about 8 dB. 12 By the time it is modulated on an RF carrier and processed through
power amplifiers it has been found to be reduced to about 4-6 dB. As a hybrid signal, when the
digital is combined with the continuous wave analog FM signal, the envelope PAPR is between 1
and 2 dB.
There are numerous topologies for creating hybrid IBOC signals, particularly with FM IBOC.
These involve various methods of generating and combining the digital and analog signals. On
the AM IBOC side, hybrid signal generation is done with few variations in technique, but the
high bandwidth-to-frequency ratio of the hybrid AM IBOC signal challenges the designer and
operator with impedance matching and bandwidth issues. These can affect transmitter loading,
system frequency and phase response, and pattern bandwidth.
The spectrum analyzer has been the instrument of choice for observing the hybrid IBOC system
passband and for seeking potential spurious emissions. However, experience has shown that
there can be differences in results obtained 1) with different analyzers, or 2) by different
operators, or 3) at different signal sampling points, when examining the same signal. While most
of these differences are minor, there have been situations where the differences result in
opposing conclusions – a system’s acceptance hangs in the balance as the conflicting results
make it uncertain whether an installation passes or fails its RF mask compliance test.
The NRSC-5 standard, as amended from time to time, specifies a general technique for
measuring the IBOC spectrum with a swept spectrum analyzer. 13 The technique involves taking
an average of a series of spectrum sweeps that requires at least 30 seconds of data collection and
at least 100 sweeps. The results of such averaging are dependent on how the selected spectrum
analyzer works, on the settings that are chosen by the operator and on the location and method of
sampling the signal. The 30-seconds/100-sweeps trace averaging method is a top-level
description that is operational in nature. It presumes the test instrument is some type of swept
analyzer and specifies three basic expectations for data collection. There is a trade-off between
sweep rate, resolution bandwidth and frequency span that may challenge the dual requirement of
11
Since the publication of this paper, DTS Inc. has acquired iBiquity Digital and is the owner and developer of HD
Radio™ technology.
12
The IBOC Handbook, David Maxson, NAB/Focal Press 2007
13
NRSC-5 Normative Reference Document #8, AM Transmission System Specifications, and #6, FM Transmission
System Specifications.
Page 49
NRSC-G201-B
30 seconds and 100 sweeps in many circumstances. Further, analyzers may employ different
methodologies that do not fit the traditional 30 seconds/100 sweeps model of data collection.
For instance, a single long-term trace can be taken on some instruments, avoiding the need to
average multiple traces, or an FFT analyzer might attack the measurement requirement
differently than a swept analyzer.
1.1.
Goals of this Paper
An emission specification must ensure that the IBOC transmission measurement meets its
conflicting goals of accuracy, repeatability, cost and efficiency. A useful goal of a measurement
specification is that it be “mathematically tractable.” A tractable specification is one that is
defined in a way that its measurement can be made without dependence on the particular
characteristics of a measuring device. This goal can be contrasted with an “operational
definition.” An operational definition might describe how to configure a particular spectrum
analyzer and evaluate the results against a measurement specification.
This paper presents a suggested tractable specification that supports the 30 second/100 sweep
operational specification in the NRSC-5 transmission system specifications and that also leaves
the opportunity for new instrumentation methods to be developed. The Guideline G-201 to
which this paper is appended, presents the method that is operationally closest to the iBiquity
specification adopted as reference documents in NRSC-5-B. This paper also provides further
guidance on using spectrum analyzers and on a preferred means of sampling the IBOC signal.
Page 50
NRSC-G201-B
2. Tractable and Suggested Operational Specifications for
Hybrid IBOC Spectral Occupancy Measurements
In pure mathematical terms, power spectral density is often presented in the basic units of
dBc/Hz. Such specifications describe pure power spectral densities per unit bandwidth. The unit
bandwidth in this canonical form is not the same as a resolution bandwidth. It is an ideal
bandwidth. Since resolution bandwidth filters are imperfect because they do not perfectly
include the specified bandwidth and exclude all other frequencies, they present some error with
respect to the ideal.
The importance of a tractable specification is that it is independent of any specific measurement
instrument. The errors and uncertainties introduced by any specific instrument’s filters can be
described, linking the measurement results back to the tractable specification. On the contrary,
when there is no tractable specification, one must assume that a specific instrument or instrument
design is the reference against which all other measurement methods must be compared. The
IBOC transmission system specifications do not describe a particular instrument architecture,
thereby leaving to chance the variations in any engineer’s measurement results.
2.1.
Hybrid FM digital signal flatness and out-of band
emissions: Specification 1— Tractable Definition
The power spectral density of the digital components of the hybrid FM IBOC signal and any out
of band and spurious emissions should not exceed the RF mask in Figure 1. As expressed in
Figure 1, in units of dBc per Hz, the mask is based on the ideal 1 Hz passband. No other filter
characteristics are incorporated in this description. The reference level (0 dBc) for the mask
shall be the power of the unmodulated analog FM carrier.
Frequency Offset Relative to Carrier
Power Spectral Density, dBc/Hz
100-200 kHz offset
-70
200-250kHz offset
[-91.4 – (|offset frequency kHz|-200) ·0.260]
250-540 kHz offset
-104.4
540-600 kHz offset
[-104.4 – (|offset frequency kHz|-540) ·0.093]
>600 kHz offset
-110
Table 1
Tractable Specification for Hybrid FM IBOC Power Spectral Density
In Decibels with Respect to Analog Carrier Power per Hertz of Bandwidth
Page 51
NRSC-G201-B
Frequency Offset, kHz
Figure 1
Tractable Specification for Hybrid FM IBOC Power Spectral Density
In Decibels with Respect to Analog Carrier Power per Hertz of Bandwidth
The tractable power spectral density mask may be scaled to compensate for the instrumentation
employed, including such instrumentation characteristics as resolution bandwidth, RBW filter
noise bandwidth, detector averaging error, and the like.
Obtaining adequate flatness across the 400 kHz passband of the hybrid FM IBOC signal is not
typically a concern, due to the ability to maintain appropriate filter and antenna bandwidths at
FM frequencies. The gain flatness of the radiated hybrid FM IBOC signal should be within ±0.5
dB, according to the standard. The group delay flatness is expected to be within 600 ns
differential across the 400 kHz bandwidth. Group delay may be assumed to be in compliance
when the group delay performance for all narrow band components (e.g. filters) in the
transmission chain collectively meets this requirement.
2.2.
Operational Variations: Bandwidth
When employing actual measurement and computational methods to determine compliance with
this mask, it will be acceptable to employ bandwidths wider than the normalized 1 Hz bandwidth
Page 52
NRSC-G201-B
and to employ filter passband characteristics not matching the ideal, as long as the mask
amplitude is adjusted accordingly.
For instance, employing a 1 kHz Resolution Bandwidth on a spectrum analyzer requires two
adjustments to the tractable mask. First, the mask in Figure 1 is shifted +30 dB to compensate
for the change from the tractable definition’s ideal 1 Hz power spectral density to 1 kHz. (This
results in the 1 kHz PSD mask adopted in NRSC-5.) Second, the mask should be shifted by an
appropriate amount to compensate for the noise bandwidth of the actual filter employed by the
analyzer.
An analog four-pole synchronous filter has a typical noise bandwidth that is 0.52 dB greater than
the nominal (-3 dB) bandwidth of the filter. The increased bandwidth causes the four-pole
filtered instrument to overstate the measured level of a noise-like signal by 0.52 dB. Employing
a four-pole filter to measure IBOC signals requires an adjustment to the mask to account for the
overstatement caused by the filtering technique. Hence, a measurement employing a 1 kHz
four-pole filter requires a 30 + 0.52 = +30.52 dB adjustment to the mask. The uncertainty of the
analog RBW will usually be 10 to 20%, thus giving an amplitude uncertainty of ±0.4 to ±0.8 dB
A Gaussian filter requires an adjustment on the order of +0.24 dB; with digitally implemented
Gaussian filters, the uncertainty in the bandwidth is usually 0.5 to 1%, thus giving an amplitude
uncertainty of ±0.02 to ±0.04 dB. See instrument specifications for the necessary filter
corrections for noise-like signal measurements.
In addition to the filter noise bandwidth issue, the detection method can affect the result. If the
swept spectrum analyzer lacks state-of-the-art digital detection and processing capability (the
ability to average on a power scale, instead of a decibel scale), it will be necessary to decrease
the mask by 2.51 dB to compensate for the analyzer’s averaging error (e.g. +30.52 dB – 2.51 dB
for 1 kHz RBW with conventional 4-pole filter and averaging error). Some older spectrum
analyzers with “Sample” detectors have this error. However, other spectrum analyzers that have
Sample detectors perform the averaging in a fashion that avoids the averaging error. The
instrument operator must determine whether the instrument presents results with or without the
error. Spectrum analyzers with “RMS” or “Average Power” detector capabilities are not
expected to have this error.
In a swept analyzer, filter bandwidths greater than 1 kHz may be too wide in some circumstances
to establish that an otherwise compliant signal is indeed in compliance with the FM IBOC mask,
due to the spreading impact of filter bandwidths on the slopes of the mask. See further
discussion in the section below titled Further Discussion of FM IBOC Measurements with
Spectrum Analyzers.
2.3.
Operational Variations: Reference Level
The reference method for establishing the analog carrier power level is with no modulation.
With care, the analog carrier power can be measured reliably with modulation present. Attention
should be paid to the manner in which a spectrum analyzer or other instrument integrates the
Page 53
NRSC-G201-B
total power of the modulated carrier; any corrections should be incorporated for the differences
between such a measurement and measuring an unmodulated, continuous wave carrier.
Attention should also be given to the linear range of the spectrum analyzer input. All RF energy
present at the input of the analyzer should total less than the 1 dB compression level of the
instrument. If there are other signals on the spectrum (such as on a multi-station transmission
line) it is the total power of all the signals that must not overload the instrument.
2.4.
Operational Variations: Averaging Time
The tractable specification does not indicate an averaging time. Instead, one should sample the
spectrum for a duration that is sufficient to minimize variations between sets of measurements
that may be made at a later time. The minimum averaging time is dependent on the
instrumentation and the span of frequencies being evaluated. To ensure sufficient averaging time,
experience has shown that capturing at least 33 symbols per unit bandwidth results in relatively
smooth traces. For less variance between measurements of the same signal, a longer acquisition
time may be applied.
For example, a swept analyzer in power detection mode conducting a single sweep across 500
kHz of spectrum for 50 seconds crosses 1 kHz in 0.1 seconds. At the FM IBOC symbol rate of
344.5 Sy/s, the analyzer dwells for 34.45 symbols’ duration in 1 kHz of spectrum, which is at
least the suggested minimum 33 symbols per unit bandwidth.
In an alternative example, repeatedly sweeping a swept analyzer in power detection mode across
the 500 kHz bandwidth at, say, 1 second per sweep 14 yields 0.69 symbols per kHz per sweep.
Multiplied by 100 sweeps, the total symbol time per kHz of spectrum is 69 symbols, well above
the recommended minimum.
On the other hand, an FFT analyzer that samples the entire frequency span of interest might
provide repeatable results with a brief continuous digital sample of the RF signal. Sampling 33
successive symbols in this fashion would take about 1/10 second (Experimentation with this type
of sampling suggests that the variance among successive measurements can be further reduced
by increasing sampling time to 4/10 to 5/10 second.).
Other methods may be devised to achieve low variance between repeated measurements.
This guidance in sweep/averaging time can be employed with older spectrum analyzers, subject
to correction of certain inherent errors or external computational analysis, or with newer
spectrum analyzers with state-of-the-art power measurement capabilities, or with dedicated
hardware designed for the purpose.
14
Operators are also advised to maintain the proper sweep rate for the selected resolution bandwidth. Using the
minimum rule of thumb, the sweep rate (Hz/s) should be no faster than ½ times the square of the resolution
bandwidth. Thus, a 500 kHz span should be swept at 1 kHz resolution bandwidth no faster than 1 second per sweep.
Spectrum analyzers have auto-coupled sweep rates to ensure this criterion is met, so long as they are not disabled or
their warnings ignored.
Page 54
NRSC-G201-B
It is anticipated that application-specific devices may employ FFT techniques to develop mask
measurements and signal statistics. The tractable specification enables instrument designers to
employ the methods that best balance efficiency, accuracy, and repeatability without reliance on
fixed bandwidths and “sweep” rates.
2.5.
Measuring Hybrid FM IBOC Spectral Occupancy against
the Mask:
The measurement of the spectral components of a Hybrid FM IBOC signal and potential out of
band and spurious emissions is begun with the selection of a measurement device to make the
measurement. The device should have low enough noise and high enough dynamic range to be
able to discriminate signal and spurious emissions from the noise floor. This requires that the
instrumentation noise floor be, for instance, less than -80 dBc/kHz for FM. It is recommended
that the average noise floor be at least ten dB below the minimum level of the mask (e.g. -90
dBc/kHz for FM measured with a 1 kHz RBW). If the instrument noise floor is too high, it can
add to a low-level spurious emission and make a mask-compliant emission appear noncompliant. Also, there should be some headroom between the total input power (average) of all
RF energy on the input port and the 1 dB compression point of the instrument. If possible, allow
10 dB between the total input power to the first stage of the instrument (after passive attenuation)
and the 1 dB compression level to insure that peaks, input level variations and minor calibration
errors do not drive the input into compression.
Once the measurement device is selected, and if it has an adjustable resolution bandwidth, a
desired RBW should be selected. The RBW should be no wider than the widest that is
appropriate for the type of filter employed by the instrument. Wider RBWs permit faster data
acquisition of the desired spectrum (such as sweeps of a swept analyzer), minimizing time spent
on each measurement without compromising accuracy. Typical four-pole filters must be limited
to 3 kHz RBW or less to keep within the FM mask slopes. Gaussian filters typically may be
employed at up to 6.8 kHz RBW. These limits are based on how the slopes of these filters
interact with the slopes of the hybrid FM IBOC RF mask.
•
The following adjustments to the mask may be required depending on the nature of the
instrument selected:
The mask in this paper is presented in dBc/Hz. Raise the mask (or deduct from the
measured value) to account for the actual bandwidth of the displayed result before
comparing results with the mask. The conversion from dBc/Hz to dBc/kHz is 30 dB. For
dBc/3 kHz to dBc/Hz, the conversion is 34.8 dB. If the instrument can report results in
dBc/Hz (literally with 1 Hz RBW filtering, or by mathematical conversion when using
other filter bandwidths), then the mask need not be adjusted to compensate for the chosen
bandwidth.
Raise the mask (or deduct from the measurement) 0.52 dB for the noise bandwidth of
four-pole synchronously tuned RBW filters, or 0.24 dB for Gaussian filters, or the
number of dB specified by the manufacturer to adjust for the noise bandwidth of the filter
Page 55
NRSC-G201-B
in use. If the manufacturer indicates that the instrument makes the filter noise bandwidth
correction internally, then do not manually make an adjustment.
•
Set input levels and for less than the nominal compression level of the instrument. Keep in
mind the presence of other strong signals on the signal source that might contribute to the
instrument’s compression of the signal under test.
•
Set the reference level by measuring the FM carrier level. The reference method is to cut
modulation, however upon verification of results it may suffice to use a wide (100 kHz, 300
kHz or 1 MHz RBW) filter, and averaging if needed, to ascertain the reference level.
On an instrument that has no true power detection feature (such as older spectrum
analyzers with a sample detector that requires trace averaging to measure digital signals)
there may be a discrepancy of up to 2.5 dB between the displayed and the actual IBOC
sideband power. In such cases, sample detection with log display and trace averaging of
the digital signal can understate the digital sideband power by up to 2.5 dB while the
analog continuous wave reference signal is measured without such a bias. Caution, some
analyzers with sample detection and trace averaging make the necessary corrections by
not averaging the traces but averaging the underlying data, in which case this error is not
present. The operator must be familiar with the instrument’s averaging method to make
the correct adjustments.
•
Set the desired display span, if the instrument permits. Ensure the display span is narrow
enough to maintain frequency accuracy.
•
Establish the data accumulation time. Ensure that the total time spent in the working
bandwidth (e.g. RBW) or the data sampling time (e.g. for an FFT analysis) amounts to at
least 33 symbols (approx 0.1 seconds) per unit bandwidth. More may be necessary to
achieve the desired reduction in variance. Multiply this by the span to obtain the sweep time,
or other data accumulation time, required.
Conduct one or more measurements of the signal.
2.6.
Measuring Hybrid FM digital signal subcarrier group power
(i.e. Primary Main, Primary Extended upper or lower sidebands):
The measurement of subcarrier group power is a means to quickly determine that the ratios
between analog power and the injected power of the various IBOC subcarrier groups is correct.
This measurement has no independent tractable component because it is derived from the
tractable specification for hybrid FM operation (Specification 1) and from the Amplitude Scale
Factors of the subcarriers. It is presented to provide guidance on measuring total power within
the hybrid FM IBOC digital sidebands.
Page 56
NRSC-G201-B
Hybrid Mode
MP1†
MP2
Primary
Main +
Primary
Extended 1
Primary
Subcarrier groups
Main (PM)
Passband start (Hz
from channel center)
Passband (Hz from
channel center)
Total number of
subcarriers
(incl. ref. if applicable)
Nominal power per
sideband (dBc)
dB referenced to PM
power (MP1 mode)
Ratio of total power
relative to PM power
MP3
Primary
Main +
Primary
Extended
1& 2
MP11
Primary
Main +
Primary
Extended
1, 2, 3 & 4
129179
122275
115371
101563
198583
198583
198583
198583
191
210
229
267
-23.0
-22.6
-22.2
-21.5
0.0
+0.4
+0.8
+1.5
1.0
1.1
1.2
1.4
†Assumes -20 dBc licensed PM total power
Table 2
Passbands and Nominal Levels of Hybrid FM IBOC Primary Main, and Primary
Extended OFDM Subcarrier Groups
The combined power of the Primary Main (“PM”) hybrid digital subcarrier sidebands is expected
to be nominally ≤-20 dBc (or -23 dBc per sideband). For additional Primary Extended subcarrier
partitions, the total computed nominal power level is the product of the subcarrier power level
(-45.8 dBc) times the total number of subcarriers, as listed in Table 2. Actual measurements
should be adjusted to compensate for the differences between the measurement method and the
ideal values provided in Table 2.
Example— Spectrum Analyzer “Channel Power” Measurement of Hybrid FM IBOC Digital
Subcarrier Groups.
Many swept spectrum analyzers have a feature that measures the power between two set-points.
This is commonly called the Channel Power function. 15 In setting up a Channel Power
measurement, there are typically two bandwidth settings to consider.
First, set the Channel Power function to look at the correct “channel.” To measure total Primary
Main and Primary Extended subcarrier groups (when operating in MP11 mode), the channel
bandwidth to be measured is nominally 100 kHz. Set a 100 kHz “channel” bandwidth. Center
15
Some instruments also have an adjacent channel power function, which is an extension of the channel power
function, the use of which is not addressed in this example.
Page 57
NRSC-G201-B
one “channel” measurement at -150 kHz (lower sideband) and a second one at +150 kHz (upper
sideband) relative to the analog carrier. 16 This creates two Channel Power measurement
“channels” at -200 to -100 kHz and +100 to +200 kHz offsets from analog center frequency.
Some instruments will internally create a virtual RBW of the selected bandwidth (100 kHz)
automatically, and no RBW setting will be necessary. Some instruments do not do as much
mathematical heavy lifting, and require the operator to employ an available resolution bandwidth
setting. It is safe to select a 1 kHz RBW for Channel Power measurements. Because a Channel
Power measurement represents total sideband power, and is not a power spectral density
measurement against the mask, the resolution bandwidth employed to make a Channel Power
measurement does not need to be 1 kHz. For a given instrument and channel noise condition,
there may be another Resolution Bandwidth setting that optimizes the efficiency (time) of the
measurement against the error produced measuring noise outside the channel bandwidth of
interest, which can be found by experimentation.
To simplify taking channel power measurements on the upper and lower sidebands separately, it
is convenient to set the span of the full sweep to 400 kHz centered on the analog carrier
frequency. This allows two Channel Power windows to be placed on the display
simultaneously - nominally from -200 to -100 kHz offsets and from +100 to +200 kHz offsets. 17
This arrangement will also conform to the stability of typical instruments. For an analyzer with
0.25%-of-span “frequency readout accuracy,” the 400 kHz span will still allow for center
frequency errors of up to 1 kHz without affecting the measurement. Thus one sweep and two
settings of channel markers can make the channel power measurement conveniently and with the
required accuracy. If the analog signal must be present for the measurement, be certain analog
modulation does not corrupt the measurement; keep analog sidebands from dominating the -3
to -30 dB slope of the RBW filter or substantially overlapping into the passband of the
measurement.
2.7.
Further Discussion of FM IBOC Measurements with
Spectrum Analyzers
Numerical modeling 18 for older spectrum analyzers with 4-pole sync-tuned filters shows that the
response of a channel bandwidth computation has 30 dB rejection at an offset 2.079 RBWs from
the edge of the channel. Thus, a 3 kHz RBW will have at least 30 dB rejection to show mask
conformance. In contrast, a 10 kHz RBW may not.
The four-pole filter passband flatness is -3 dB at the nominal bandwidth and within 0.5 dB from
0.589 RBWs inward.
16
The round number of 100 kHz has been shown experimentally to be effective. To be more precise, the channel
start and stop frequency offsets for operation in MP1, MP2, MP3 and MP11 modes are shown in Table 2.
Instruments may be set to these precise frequencies if they have the capability.
17
If the instrument can perform only one channel power measurement at a time, then set it up for one sideband and
perform the measurement; and set up for the other sideband and measure. Since the selected span covers both
sidebands, no readjustment of the span is necessary between the two Channel Power measurements.
18
Acknowledgement is given to Mr. Joe Gorin of Agilent for providing analysis of spectrum analyzer filter
performance and other valuable assistance in generating this document.
Page 58
NRSC-G201-B
For Gaussian RBW filters, the response of a channel bandwidth computation has 30 dB rejection
at an offset 1.312 RBWs from the edge of the channel. Thus, a 6.8 kHz RBW meets the 30 dB
rejection objective at the sides of the passband while still allowing a margin for frequency
accuracy of 1 kHz.
The passband flatness of a Gaussian RBW filter is within 0.5 dB from 0.524 RBWs inward.
Page 59
NRSC-G201-B
3. AM
In pure mathematical terms, power spectral density is often presented in the basic units of
dBc/Hz. Such specifications describe pure power spectral densities per unit bandwidth. The unit
bandwidth in this canonical form is not the same as a resolution bandwidth. It is an ideal
bandwidth. Since resolution bandwidth filters are imperfect because they do not perfectly
include the specified bandwidth and exclude all other frequencies, they present some error with
respect to the ideal.
The importance of a tractable specification is that it is independent of any specific measurement
instrument. The errors and uncertainties introduced by any specific instrument’s filters can be
described, linking the measurement results back to the tractable specification. On the contrary,
when there is no tractable specification, one must assume that a specific instrument or instrument
design is the reference against which all other measurement methods must be compared. The
IBOC transmission system specifications do not describe a particular instrument architecture,
thereby leaving to chance the variations in any engineer’s measurement results.
3.1.
Hybrid AM digital signal flatness and out-of-band
emissions: Specification 2 – Tractable Definition
The power spectral density of the digital components of the hybrid AM IBOC signal and any out
of band and spurious emissions should not exceed the RF mask in Figure 2. As expressed in
Figure 2, in units of dBc per Hz, the mask is based on the ideal 1 Hz passband. No other filter
characteristics are incorporated in this description. The reference level (0 dBc) for the mask
shall be the power of the unmodulated analog AM carrier.
Frequency Offset Relative to Carrier
5-10 kHz offset*
10-15 kHz offset
15-15.2 kHz offset
15.2-15.8 kHz offset
15.8-25 kHz offset
25-30.5 kHz offset
30.5-75 kHz offset
>75 kHz offset
Power Spectral Density, dBc/Hz
-59.1
-51.6
-52.8
-63.8 – (|offset frequency kHz|-15.2) ·43.3
-89.8
-89.8 – (|offset frequency kHz|-25) ·1.273
-96.8 – (|offset frequency kHz|-30.5) ·0.292
-109.8
Table 3
Tractable Specification for Hybrid AM IBOC Power Spectral Density
In Decibels with Respect to Analog Carrier Power per Hertz of Bandwidth
*This table assumes AM analog operation at 5 kHz audio bandwidth; for the 8 kHz option, this entry changes to “810 kHz offset” to accommodate the presence of analog sideband energy between 5 and 8 kHz offsets.
Page 60
NRSC-G201-B
Figure 2
Tractable Specification for Hybrid AM IBOC Power Spectral Density
In Decibels with Respect to Analog Carrier Power per Hertz of Bandwidth
Further, it is recommended that the Hybrid AM IBOC signal be transmitted with an amplitude
flatness ±0.5 dB to 10 kHz off center and ±1 dB to 15 kHz. 19 NRSC-5-C calls for phase
response to be within ±3 µs across the 30 kHz spectrum of the signal 20, with the antenna system
optimized for Hermitian symmetry. Recognizing that some AM antenna systems may not be
able to fully comply with this flatness criterion, this flatness specification is advisory.
3.2.
Operational Variations: Bandwidth
When employing actual measurement and computational methods to determine compliance with
this mask, it will be acceptable to employ bandwidths wider than the normalized 1 Hz bandwidth
and to employ passband characteristics not matching the ideal, as long as the mask amplitude is
adjusted accordingly.
For instance, employing a 300 Hz resolution bandwidth on a spectrum analyzer requires two
adjustments to the mask. First, the tractable mask is shifted +24.8 dB to compensate for the
change from the tractable definition’s 1 Hz power spectral density to 300 Hz. Second, if
applicable, the mask should be shifted by an appropriate amount to compensate for the noise
bandwidth of the actual filter employed by the analyzer. A traditional physical four-pole
synchronous filter has a typical noise bandwidth that is 0.52 dB greater than the ideal bandwidth
19
NRSC-5-B specifies these figures. On the contrary iBiquity suggests in the NAB Engineering Handbook, 10th
edition, that the specification be ±0.5 dB to 5 kHz and ±4 dB to 15 kHz.
20
NRSC-5-B specifies this. In the NAB Engineering Handbook, 10th edition, the suggested figure is 5 µs.
Page 61
NRSC-G201-B
of the same nominal value. The increased bandwidth causes the four-pole filtered instrument to
overstate the measured level of a noise-like signal by 0.52 dB. Employing a four-pole filter to
measure IBOC signals requires an adjustment to the mask to account for the overstatement
caused by the filtering technique. Hence, a measurement employing a 300 Hz four-pole filter
requires a 24.8 + 0.52 = +25.4 dB adjustment to the Figure 2 mask.
A Gaussian filter would require a correction of +0.27 dB; typical digitally implemented
approximations to a Gaussian filter require no correction.
In a swept analyzer, 4-pole filter bandwidths greater than 300 Hz may be too wide to establish
that an otherwise compliant signal is indeed in compliance with the mask, due to the spreading
impact of filter bandwidth on the slopes of the mask. See further discussion in the section below
titled Further Discussion of AM IBOC Measurements with Spectrum Analyzers.
3.3.
Operational Variations: Reference Level
The reference method for establishing the analog carrier power level is with no modulation.
With care, the analog carrier power can be measured with modulation present. Even with
asymmetrically modulated AM signals (e.g., +125/-99%) the average power of the carrier, as
measured with a narrow RBW (e.g. ≤300 Hz) is typically a reliable indicator of the carrier power
reference. Attention should be paid to the manner in which a spectrum analyzer or other
instrument integrates the total power of the modulated carrier; any corrections should be
incorporated for the differences between such a measurement and measuring an unmodulated,
continuous wave carrier.
3.4.
Operational Variations: Averaging Time
The tractable specification does not indicate an averaging time. Instead, one should sample the
spectrum for a duration that is sufficient to minimize variations in successive measurements.
The minimum averaging time is dependent on the instrumentation and the span of frequencies
being evaluated. To ensure sufficient averaging time, experience has shown that capturing at
least 33 symbols per unit bandwidth results in relatively smooth traces. For less variance
between measurements of the same signal, a longer acquisition time may be applied.
For example, a swept analyzer in power detection mode sweeping 40 kHz of spectrum once for
30 seconds crosses 300 Hz in 0.23 seconds. At a symbol rate of 172.3 Sy/s, the analyzer dwells
for 39 symbols’ duration in 300 Hz of spectrum, which is at least 33 symbols, as recommended.
In an alternative example, repeatedly sweeping a swept analyzer in power detection mode across
the 40 kHz bandwidth at, say, 1 second per sweep 21, yields 1.3 symbols per 300 Hz per sweep.
Multiplied by 30 sweeps, the total symbol time per 300 Hz of spectrum is 39 symbols.
21
Operators are also advised to maintain the proper sweep rate for the selected resolution bandwidth. Using the
minimum rule of thumb, the sweep rate (Hz/s) should be no faster than ½ times the square of the resolution
bandwidth. Thus, a 45 kHz span should be swept at 300 Hz resolution bandwidth no faster than 1 second per sweep.
Spectrum analyzers have auto-coupled sweep rates to ensure this criterion is met, so long as they are not disabled or
ignored.
Page 62
NRSC-G201-B
On the other hand, an FFT analyzer that samples the entire frequency span of interest might
provide repeatable results with a brief continuous digital sample of the RF signal. Sampling 33
successive symbols in this fashion would take about 1/5 second (Experimentation with this type
of sampling suggests that the variance among successive measurements can be further reduced
by increasing sampling time to 4/5 to 1 second.).
Other methods may be devised to achieve low variance between repeated measurements.
This guidance can be employed with older spectrum analyzers, subject to correction of certain
inherent errors or external computational analysis, or with newer spectrum analyzers with stateof-the-art power measurement capabilities, or with dedicated hardware designed for the purpose.
It is anticipated that application-specific devices may employ FFT techniques to develop mask
measurements and signal statistics. The tractable specification enables instrument designers to
employ the methods that best balance efficiency, accuracy, and repeatability without reliance on
fixed bandwidths and “sweep” rates.
3.5.
Measuring Hybrid AM IBOC Spectral Occupancy against
the Mask:
The measurement of the spectral components of a Hybrid FM IBOC signal and potential out of
band and spurious emissions is begun with the selection of a measurement device to make the
measurement. The device should have low enough noise and high enough dynamic range to be
able to discriminate signal and spurious emissions from the noise floor. This requires that the
instrumentation noise floor be, for instance, less than -85 dBc/300 Hz for AM. It is
recommended that the average noise floor be at least ten dB below the minimum level of the
mask (e.g. -95 dBc/300 Hz for AM measured with a 300 Hz RBW). If the noise floor is too
high, it can add to a low-level spurious emission and make a mask-compliant emission appear
non-compliant.
Once the measurement device is selected and if it has an adjustable resolution bandwidth, select
the widest appropriate RBW, based on the type of filter employed by the instrument. Wider
RBW’s permit faster data acquisition of the desired spectrum (such as sweeps of a swept
analyzer), minimizing time spent on each measurement without compromising accuracy.
Typical four-pole filters must be limited to 300 Hz RBW’s or less. Gaussian filters may be
employed at 1 kHz RBW.
•
The following adjustments to the mask may be required depending on the nature of the
instrument selected:
The mask is presented in dBc/Hz. Raise the mask (or deduct from the measured value) to
account for the actual bandwidth of the displayed result before comparing results with the
mask. The conversion from dBc/Hz to dBc/kHz is 30 dB. For dBc/300 Hz the
conversion is 24.8 dB. If the instrument can report results in dBc/Hz, then the mask need
not be adjusted to compensate for the chosen bandwidth.
Page 63
NRSC-G201-B
Raise the mask (or deduct from the measurement) 0.52 dB for four-pole synchronously
tuned RBW filters, or 0.24 dB for Gaussian filters, or the number of dB specified by the
manufacturer to adjust for the noise bandwidth of the filter in use. If the manufacturer
indicates that the instrument makes the filter noise bandwidth correction internally, then
do not manually make an adjustment.
•
Set input levels and for less than the nominal compression level of the instrument. Keep in
mind the presence of other strong signals on the signal source that might contribute to the
instrument’s compression of the signal under test.
•
Set the reference level by measuring the AM carrier level. The reference method is to cut
modulation, however upon verification of results it may suffice to use a narrow filter and
averaging to ascertain the reference level.
On an instrument that has no true power detection feature (such as older spectrum
analyzers with a sample detector that requires trace averaging to measure digital signals)
there may be a discrepancy of up to 2.5 dB between the use of an unmodulated carrier
reference and the use of trace averaging to measure digital signal power spectral density.
In such cases, this type of detection and averaging of the digital signal understates the
digital power on a logarithmic detector by up to 2.5 dB while the analog Continuous
Wave reference signal is measured without such a bias.
•
Set the desired display span, if the instrument permits. Ensure the display span is narrow
enough to maintain frequency accuracy.
•
Establish the data accumulation time. Ensure that the total time spent in the working
bandwidth (e.g. RBW) or the data sampling time (e.g. for an FFT analysis) amounts to at
least 33 symbols (approx 0.2 seconds) per unit bandwidth, or more if necessary to reduce
measurement variance. Multiply this by the span to obtain the sweep, or other data
accumulation, time required.
Conduct one or more measurements of the signal.
Page 64
NRSC-G201-B
3.6.
Measuring Hybrid AM digital signal subcarrier group power
(i.e. Primary, Secondary, and Tertiary upper or lower
sidebands):
The measurement of subcarrier group power is a means to quickly determine that the ratios
between analog power and the injected power of the various IBOC subcarrier groups is correct.
The measurement of subcarrier group power has no independent tractable component because it
is derived from the tractable specification for hybrid AM operation (Specification 2) and from
the Hybrid AM IBOC OFDM amplitude scale factors in the NRSC-5 Reference Documents. It is
presented to provide guidance on measuring total power within each of the hybrid AM IBOC
digital sidebands.
Recommended maximum power levels, based on Specification 2, are presented in Table 4.
Actual measurements should be adjusted to compensate for the differences between the
measurement method and the ideal values provided in Table 4.
AM IBOC Hybrid Mode
Subcarrier groups
Passband Start Frequency
Offset
Passband Stop Frequency
Offset
Maximum power per
sideband (upper or lower)
Nominal power per
sideband
High Power Option
Primary
10 kHz
Secondary
5
Tertiary
0.2*
15 kHz
10
5
-15 dBc
-22.5
-29.5
-16 dBc
-23
-30
Table 4
Passbands, Nominal Levels, and Maximum Levels of Hybrid AM IBOC Primary,
Secondary, and Tertiary OFDM Subcarrier Groups
* Nominal offset to minimize impact of unmodulated analog carrier and Reference subcarriers.
**These specifications are the same for 5- and 8-kHz analog audio bandwidth systems, except that more care must
be taken to limit analog AM sidebands in the 5-10 kHz offset region during measurements of the Secondary
subcarriers.
Example— Spectrum Analyzer Total Channel Power Measurement of Hybrid AM IBOC Digital
Subcarrier Groups.
There are two bandwidth settings to consider. First, set the Channel Power bandwidth to be
measured to 5 kHz. This will cover one of the OFDM subcarrier sideband groups. Center the 5
kHz Channel Power bandwidths of each measurement at −12.5, -7.5, +12.5, and +7.5 kHz
(Primary and Secondary, lower and upper sideband groups, respectively) relative to the carrier.
The Tertiary sidebands reside beside the AM carrier and beneath the AM modulation, so a total
power measurement of the Tertiary sidebands’ energy is not obtained with a typical spectrum
Page 65
NRSC-G201-B
analyzer unless modulation is removed. Then set the optimum Resolution Bandwidth for the
instrument to employ to integrate across the Channel Power bandwidth. To obtain a
measurement of all the subcarriers in one sideband with Gaussian RBWs, as noted above, the
RBW can be set as wide as 1 kHz, or narrower. With the four-pole synchronously tuned filters,
the widest RBW that will enable a Channel Power measurement that will not fail on the mask
slope is 300 Hz RBW.
3.7.
Further Discussion of AM IBOC Measurements with
Spectrum Analyzers
Numerical modeling for older spectrum analyzers with 4-pole sync-tuned filters shows that the
response of a channel bandwidth computation has 30 dB rejection at an offset 2.079 RBWs from
the edge of the channel. Thus, a 300 Hz RBW will have at least 30 dB rejection necessary to
show mask conformance. In contrast, a 4-pole 1 kHz RBW filter will not. This is a particularly
challenging situation due to the relative steepness of the Hybrid AM IBOC mask slopes,
compared to the FM Hybrid IBOC mask. The AM mask is just useable with four-pole filters
when scaled up to 300 Hz power spectral density because the slope outside the OFDM subcarrier
frequencies approximates that of a 300 Hz four-pole synchronously tuned spectrum analyzer
filter measuring an ideally modulated Primary sideband. 22
The four-pole filter passband flatness is -3 dB at the nominal bandwidth and within 0.5 dB from
0.589 RBWs inward. Thus, a 300 Hz RBW would meet the passband flatness requirement from
10.18 kHz to 14.82 kHz.
For Gaussian RBW filters, the response of a channel bandwidth computation has 30 dB rejection
at an offset 1.312 RBWs from the edge of the channel. Thus, a Gaussian 1 kHz Hz RBW meets
the rejection objective of the mask at the sides of the OFDM subcarrier passband.
The passband flatness of a Gaussian RBW filter is within 0.5 dB from 0.524 RBWs inward.
Thus, a 1 kHz Gaussian RBW would meet the passband flatness requirement from 10.53 kHz to
14.47 kHz. With Gaussian RBWs, the RBW can be set to 1 kHz or narrower, while still allowing
for frequency errors of 188 Hz. For an analyzer with 0.25%-of-span “frequency readout
accuracy,” the span may be up to 75 kHz for this measurement, thus one sweep and two
positionings of a “band marker” can make the Primary Sideband total power measurements
conveniently and with the required accuracy.
The process may be repeated at -7.5 and +7.5 kHz offsets relative to the carrier to obtain the
Secondary Sideband measurements from -5 to -10 kHz and +5 to +10 kHz. Be certain analog
modulation does not corrupt the measurement; keep analog sidebands from dominating the -3 to
-30 dB slope of the RBW filter or substantially overlapping into the passband of the
measurement.
22
Maxson, D., The Role of The Detector in Spectrum Analyzer Measurement of Digital Signals, National
Association of Broadcasters 2008 Broadcast Engineering Conference Proceedings
Page 66
NRSC-G201-B
4. Discussion of Spectrum Analyzer Measurements
4.1.
Making Measurements to the Standards with Newer
Spectrum Analyzers
Newer spectrum analyzers have capabilities and convenience features that allow these kinds of
measurements to be made easily and accurately. Ideally, an analyzer has a power-responding
detector or suitable work-around, and a power-integrating marker or measurement function.
The power-responding detector is often called an “RMS detector” because it responds
proportionally to the mean-square of the changing voltage of the signal, in other words,
proportionally to the power. Manufacturers also may call this detector the “average detector” or
“average detector: power” because it responds to the average of a signal parameter within a short
duration; with this detector, the user has the choice of whether to average the voltage, the log of
the voltage, or, as desired in this case, the power. Care should be taken to set the video
bandwidth (VBW) of the analyzer much wider than the resolution bandwidth (RBW); this is
necessary because the VBW filter typically operates on the display scale, which is usually set to
a log scale, and averaging on the log scale does not achieve the correct result for noise-like
signals.
A specialized “power integrating” feature of the digital spectrum analyzer is often called a
“channel power measurement” or a “band power marker.” These measure the power across some
width of the spectrum analyzer display by summing the measurement points (sometimes called
“buckets” or “pixels”) and correcting for the effects of the point spacing and the noise bandwidth
of the RBW being used.
Some modern analyzers have RBW filters with much higher selectivity than the 4-pole filters
that have traditionally been commonplace. Higher selectivity allows wider RBWs to be used
without polluting the desired measurement band with adjacent spectral energy, and alternatively
without causing a measured signal to appear to exceed a steep mask slope. The variance of a
power measurement of a noise-like signal is inversely proportional to the bandwidth, so such a
modern analyzer can make measurements proportionally faster for a constant variance in results.
4.2.
Making Measurements to the Standards with Older
Spectrum Analyzers: Exact but Inconvenient Techniques
As mentioned above, older analyzers may not have suitable detectors, suitable “channel power”
functions or markers, or high selectivity RBWs.
Page 67
NRSC-G201-B
If the older analyzer does not have channel power functions, some instruments permit the trace
data to be exported to an external computer which can be programmed to make a computation
and incorporate the necessary adjustments..
Finally, without an RMS-style detector, the best measurement technique is to use a VBW of 10
or more times the RBW. The “Sample” detector should be used with multi-sweep averaging of
the results in linear power units. This technique will allow power-scale response whether the
channel power summation is done with an internal or external computation. 23
4.3.
Making Measurements to the Standards with Older
Spectrum Analyzers: Inexact but Convenient Technique
Older analyzers do not have power-scale averaging processes, such as the RMS detector. They
average signal, whether with the VBW filter or trace averaging, on the “display scale” of the
analyzer. Usually, this scale is set to a logarithmic scale, such as 10 dB/division, but can be set to
the “linear” (linear in volts, that is) scale. Neither of these scales is linear in power.
If the signal being measured has a statistical distribution that is assured to be noise-like, the
response of the analyzer due to its averaging on the log scale is well known to be an underresponse of 2.506 dB. Therefore, heavy filtering from trace averaging or VBW filtering can be
used, and the analyzer response can be compensated by adding 2.506 dB to the result. If there is
a CW-like spurious signal dominating the measurement, though, this technique will overcorrect
for the average power of the CW signal. The error on the spurious CW signal is in the direction
that makes a device-under-test more likely to fail with such an interfering signal, and thus would
not increase the risk of passing a device that should have failed. So this error source might be an
acceptable compromise for the convenience of averaging in the older spectrum analyzer.
3.4.1. How Close to Gaussian is the IBOC Digital Signal?
It turns out to be an excellent approximation to assume that the IBOC digital signal is white
noise-like (Gaussian in its amplitude-over-time distribution). Experiments with OFDM signals
with properly scaled RBW filters shows that the error in averaging the logs of a series of samples
within a given bandwidth is nearly identical to the error that would occur with a purely Gaussian
noise signal. The log average of the IBOC signal in the 1 kHz RBW (300 Hz for AM) was 0.04
dB higher than predicted for a Gaussian signal. Thus, the user of older analyzers who employs
the 2.506 dB adjustment for noise-like modulation will see shrinkage of the margin by only 0.04
dB. In fact, if rounded to the nearest 1/10 dB, which is common in field measurements of
broadcast signals, the Gaussian error and the IBOC error in making log averages is
indistinguishable.
23
The computation is fairly straightforward: sum the powers (e.g., in mW, not dBm) of all the data points in the
passband; divide by the number of data points in the passband; divide by the effective noise bandwidth of the RBW
filter; multiply by the bandwidth of the passband. See, for example, Agilent Technologies, Inc., Application Note
1303, Spectrum Analyzer Measurements and Noise, for details on this computation.
Page 68
NRSC-G201-B
4.4.
Averaging Techniques
This Section discusses the tradeoffs between trace averaging and video filtering, and between
longer sweeps and trace averaging.
4.4.1. Trace Averaging vs. Video Filtering
Trace averaging acts to reduce the variance of spectrum analyzer results that vary on a trace-bytrace time scale. Video filtering acts to reduce the variance when the results vary within the time
scale associated with sweeping through a resolution bandwidth filter. (It is beyond the scope of
this discussion to consider the tradeoffs in FFT-based, rather than sweep-based, spectrum
analyzers, but the conclusions are similar.)
If the signal is incoherent with the sweeping, as digital modulations are always expected to be,
then the two forms of variance reduction are similarly effective. There is a subtle tradeoff
between throughput and convenience. The manual user often prefers trace averaging because an
approximation to his final result appears very early in the averaging process as he watches it
accumulate. The remote user often prefers a narrow VBW filter because the overhead of
“retrace” (starting a new sweep) occurs only once, improving throughput. By experimentation it
has been shown that use of the trace average on IBOC signals gives a crisper, more readily
interpreted result than VBW filtering. 24 It is recommended that for IBOC measurements a traceaveraged or long-period single trace measurement be employed instead of video filtering. In this
case, the VBW filter should be set to be wider than the resolution bandwidth.
This throughput and convenience tradeoff is not available to the user making “channel power”
measurements on a power scale with an external computer attached to an analyzer. If the trace is
filtered with a VBW filter operating on the decibel scale before being transmitted to an external
computer trying to achieve power response by power summing individual elements, the effect of
the VBW averaging on the decibel scale is to cause errors up to 2.51 dB. This defeats one of the
purposes of removing trace data from the analyzer for external processing— performing
computations on the power scale instead of the log scale or voltage scale.
4.4.2. Long Sweep Versus Trace Averaging
There is also a subtle tradeoff between long sweeps and trace averaging that applies when an
RMS detector is in use. As in the case with the VBW filter acting on each sweep point, the
average detector averages the signal power across the duration of each data point on the sweep
(“bucket”). As long as the signal is noise-like or otherwise not coherent with the sweeping, the
user can either spend more time on each point, or take more sweeps, and in both cases, get the
same reduction in variance.
24
David Maxson, Measuring Your IBOC Spectrum, NAB Radio Show, 2004
Page 69
NRSC-G201-B
Depending on the implementation of the manufacturer, the average power detector in multi-trace
averaging might simply average a series of traces, yielding an averaging error, albeit minor 25, or
it might, more appropriately perform a power computation on the raw data of the accumulated
series of sweeps to obtain a more accurate trace-averaged power reading. It may be difficult to
get an answer, with certainty, from a manufacturer as to what method the analyzer uses to traceaverage in RMS or power detector mode.
25
When an average detector is employed, the error in trace averaging is substantially reduced, compared to the 2.51
dB error when trace averaging with a sample detector. This is because the average detector already has taken an
average power measurement of numerous data points in each bucket, thus reducing the “noisiness” of each of the
average power detector data points that are subsequently averaged together.
Page 70
NRSC-G201-B
ANNEX 2: TEST AND MEASUREMENT EQUIPMENT SELF-CERTIFICATION LIST
Page 71
NRSC-G201-B
ANNEX 2
NRSC-G201, NRSC-5 RF Mask Compliance: Measurement Methods and Practice
Test and Measurement Equipment Self-Certification List
_UPDATED
December 30, 2016_
The test and measurement equipment listed on this page has been self-certified by the manufacturer as being suitable for determining if an RF signal is
compliant with the NRSC-5 AM IBOC and FM IBOC emission masks. The manufacturers of all equipment listed herein have completed the form included in Annex
3 of the Guideline and submitted it to the NRSC, requesting inclusion in this list. An updated list is published as new items are added.
Neither the NRSC nor its members, participants or co-sponsors make any claim as to the suitability of this equipment for use in making the RF mask
compliance measurements described in the NRSC-G201 Guideline. Parties interested in making these measurements need to verify for themselves that this
equipment is in fact suitable for their measurement needs.
Anyone wishing to provide information to the NRSC as to the suitability or unsuitability of this equipment for IBOC RF mask compliance measurements should
send an email to nrsc@nab.org. Please include in this email the manufacturer’s name and equipment model number of the unit(s) being discussed.
Manufacturer
No
Name
Website
Self-certified for:
Model
Description
1
2
3
Page 72
Date of
selfcertification
AM
IBOC
FM
IBOC
Supports
IBOC mask
limit lines?
NRSC-G201-B
ANNEX 3: TEST AND MEASUREMENT EQUIPMENT SELF-CERTIFICATION FORM
Page 73
NRSC-G201-B
NRSC-G201-B, NRSC-5 RF Mask Compliance: Measurement Methods and Practice
Test and Measurement Equipment Self-Certification Form
Please submit completed
form to:
National Association of Broadcasters
1771 N Street, N.W.
Washington, DC 20036
Attn: Science & Technology Department
Email: nrsc@nab.org
Fax: 202-775-4981
This form is for manufacturers of
test and measurement equipment
suitable for determining if an RF signal is compliant with the NRSC-5 AM IBOC and FM IBOC emission masks, in
accordance with the measurement techniques discussed in the NRSC-G201-B Guideline. Please complete this form
and submit it to NAB by mail, fax, or email, using the contact information above.
Completed forms will be reviewed by the NRSC and if found complete, the equipment described therein will be added
to Annex 2 of the NRSC-G201-B Guideline, which lists test and measurement equipment that has been self-certified
by the manufacturer as being suitable for IBOC mask compliance measurements.
Company information:
COMPANY
NATURE OF YOUR BUSINESS
ADDRESS
CITY
PHONE (MAIN NUMBER)
FAX
STATE
ZIP CODE
STATE
ZIP CODE
WEBSITE
Contact information:
NAME
TITLE
ADDRESS
CITY
PHONE
FAX
EMAIL ADDRESS
Page 74
NRSC-G201-B
Test and measurement equipment description (use a separate page 2 of this form for each model being self-certified):
MODEL NUMBER
(OPTIONAL) VERSION
SELF-CERTIFYING FOR:
ο AM IBOC
DESCRIPTION:
ο
ο SPECTRUM ANALYZER
FORM FACTOR:
ο
HAND-HELD
ο
FM IBOC
ο
MODULATION ANALYZER
PORTABLE
ο
RACK-MOUNTABLE
ο
OTHER ____________________________________________________________________
ο
OTHER ____________________________________________________________________
Is there a “Tech Note” for this model which discusses IBOC mask compliance measurements?
If YES, please provide a reference to this Tech Note (URL preferred):
ο
YES
ο
NO
Does instrument support the use of limit lines for FM, AM IBOC masks (see Section 6.1.5 of NRSC-G201-B)? If so,
please describe here:
Self-certification items –check off all that apply:
Hybrid AM IBOC mask compliance measurements (check off each item to certify compliance):
• Minimum frequency range: 400 kHz –2 MHz
• Resolution bandwidth (RBW) settable to 300 Hz
• Video bandwidth (VBW) – no VBW filter or one that is settable to at least 3 kHz
• Peak-hold measurement mode
• Averaging (RMS) detection and/or trace averaging (with sample detection) measurement mode(s) (“max/min”
averaging is not applicable under this item)
• Dynamic range sufficient to provide accurate measurement of signal under test with respect to AM IBOC mask
• Supports calibrated measurement with 300 Hz RBW, 3 kHz VBW, 200 kHz span
Hybrid FM IBOC mask compliance measurements (check off each item to certify compliance):
• Minimum frequency range: 85 MHz – 110 MHz
• Resolution bandwidth (RBW) settable to 1 kHz
• Video bandwidth (VBW) – no VBW filter or one that is settable to at least 10 kHz
• Peak-hold measurement mode
• Averaging (RMS) detection and/or trace averaging (with sample detection) measurement mode(s) (“max/min”
averaging is not applicable under this item)
• Dynamic range sufficient to provide accurate measurement of signal under test with respect to FM IBOC mask
• Supports calibrated measurement with 1 kHz RBW, 10 kHz VBW, 2 MHz span
By signing and submitting this form I certify that:
● I am authorized to make this submission on behalf of (company name)
● I understand that the inclusion of this equipment in Annex 2 of NRSC-G201-B is at the sole discretion of the NRSC
and that the NRSC may remove it from this list at any time.
SIGNATURE
DATE
Page 75
NRSC-G201-B
ANNEX 4 – A METHOD FOR MEASURING HYBRID FM IBOC SIGNALS ON TRANSMISSION
SYSTEMS WITH INDEPENDENT DIGITAL AND ANALOG TRANSMISSION LINES USING A
CHIMP (COMBINED HYBRID IBOC MEASUREMENT PACKAGE)
Page 76
NRSC-G201-B
A method for Measuring
Hybrid FM IBOC Signals
on
Transmission Systems
with
Independent Analog and Digital Signal
Transmission Lines
such as
Dual-Input-Antenna Systems and
Separate–Antenna Systems
Utilizing a CHIMP
(Combined Hybrid IBOC Measurement Package)
Randy Mullinax
Senior Vice President Engineering
Clear Channel Radio
With Editorial Assistance by
David Maxson
Broadcast Signal Lab
October 2011
Page 77
NRSC-G201-B
This document demonstrates a method for verifying that a hybrid IBOC transmission system
whose analog and digital components are maintained on separate transmission lines complies
with the NRSC-5-C standard’s RF mask. The use of dual-input antennas or separate antennas for
analog and digital transmission requires special measurement procedures to demonstrate the
compliance of a transmitted IBOC signal with the iBiquity HD Radio FM Transmission System
Specifications (NRSC-5-C Normative Reference Document 6, with further discussion in NRSC
Guideline Document G-201-A). These transmission architectures have no common transmission
line from which the combined digital and analog signals can be sampled before they reach their
antenna(s). For this reason, this paper uses the term separate line to describe any IBOC system
that maintains separate transmission lines for the analog and digital IBOC signals, such as dualinput antenna systems and separate antenna systems.
Since there is no point in a separate-line system where the combined hybrid IBOC signal can be
sampled, additional steps must be taken to assess the hybrid signal, including establishing the
analog reference level for measurement of the digital signal. Also, the combined emissions of
the two transmission systems – digital and analog – may sum to exceed the RF mask, while each
individual transmission appears compliant. This measurement procedure will not determine the
required digital transmitter power output. This value must be computed from the required digital
ERP (see NRSC-G202 “FM IBOC Total Digital Sideband Power for Various Configurations”),
the antenna gain, line loss and any other system losses. The required output power of the digital
transmitter should be computed in advance and the digital transmitter should be operating at this
power level before the measurements begin.
Equipment List
To facilitate the measurements, the following equipment is required.
QTY.
1
1
2
1
-
DESCRIPTION
Spectrum Analyzer (Agilent E4402B or equivalent)
Splitter/Combiner (MiniCircuits ZFSC-2-2+ or equivalent)
6 dB Attenuator Pads
Variable Attenuator or assortment of fixed pads
Coax Cables for hookup (double-shielded preferred)
In addition to the equipment listed above, there must be a directional coupler or other suitable
directional RF sample available in both the analog and digital transmission lines to facilitate the
measurements. An RF sample that is not directional will cause significant measurement error.
Careful attention must be paid to the specifications of the directional couplers employed in the
analog and digital transmission lines. Determine from calibration data what the errors are for
each coupler at the expected power levels and transmitter frequencies.
Connect the test equipment as shown in Figure 1. Use a 6 dB attenuator pad at each input to the
Splitter/Combiner (if the MiniCircuits ZFSC-2-2 is used, the “reject load” shown in Figure 1 is
internal). The spectrum analyzer should be pre-configured as normal for hybrid FM IBOC
measurements.
Page 78
NRSC-G201-B
CHIMP
Figure 1
Combined Hybrid IBOC Measurement Package (“CHIMP”)
Sampling FM IBOC Signals from Separate Transmission Lines
Typical Settings
Frequency – Center frequency of FM carrier
Span – 2.0 MHz
Detector – RMS
Data Points – Maximum
Resolution Bandwidth – 1 kHz
Video Bandwidth – ≥10 kHz
Sweep/Trace Averaging – ON – up to 100 sweeps
Averaging Detector Type – Pwr (also called RMS on some instruments)
Marker – Normal – Center frequency of FM carrier
Input power level to the spectrum analyzer mixer should be well below the instrument’s 1 dB
compression point. See G-201-A Section 6.1.1 for detailed information regarding acceptable
input power levels.
Maintain Stable Loads on Splitter/Combiner
The 6 dB pads should remain connected to the splitter/combiner at all times during the
measurements to avoid significant errors. These pads provide the required isolation when one of
the sample cables is removed to establish reference levels. The spectrum analyzer is assumed to
have a precision 50 Ω input impedance, which is necessary to properly load the combiner unit.
Measurements illustrated in this paper were conducted using the equipment listed while
monitoring the reference level with a short circuit, open circuit and 50 Ω load connected to the 6
dB attenuator pad on the opposite input of the splitter/combiner from that of the sample being
measured. In all cases, the difference in the reference level was found to differ by less than 0.2
Page 79
NRSC-G201-B
dB. The pads also reduce any errors associated with the reactive nature (poor source match) of
the directional RF samples.
Initial Reference Levels
To establish the “analog reference level” (the basis is the level of the unmodulated analog carrier
- Since it is often impractical to interrupt modulation, essentially the same result will be obtained
by measuring the total power of the frequency modulated wave within its occupied bandwidth)
disconnect the digital RF sample cable from the 6 dB attenuator pad and leave the pad connected
to the splitter/combiner. Change the Resolution Bandwidth to 1.0 Mhz and the Video Bandwidth
of the spectrum analyzer to 1.0 MHz or greater and restart averaging. Adjust the reference
amplitude for an on-screen display. The display should be similar to Figure 2 (only about 10
sweeps are necessary to allow the analyzer to average the analog reference level).
Figure 2
Establishing Analog FM Reference Level
with Wide Resolution Bandwidth
Make a notation of the level indicated by the marker. In this example it is +9.1 dBm.
Page 80
NRSC-G201-B
Note that the 1.0 MHz Video Bandwidth will have very little effect on the analog reference level
(less than 0.1 dB), but the higher Video Bandwidth is required to obtain a accurate digital
measurements.
To establish the “digital reference level”, reconnect the digital RF sample cable and disconnect
the analog RF sample cable from the 6 dB attenuator pad but again leave the pad connected to
the splitter/combiner. Restart averaging and if necessary, adjust the reference amplitude for an
on-screen display (restart averaging again if any adjustment to the reference amplitude is made).
Allow the analyzer to average at least 50 sweeps. The display should appear similar to Figure 3.
Figure 3
Establishing IBOC Digital Reference Level
with Wide Resolution Bandwidth
Make a notation of the level indicated by the marker which in this example is -13.2 dBm.
Page 81
NRSC-G201-B
Calculate Level Correction Required
From the RF levels noted above, calculate the amount of additional attenuation required so that
the digital reference level will be as close as possible to the digital sideband power level (in dBc)
below the analog reference level. 26 The digital sideband power level will be the same as the
“Total digital sideband ERP (dBc ref. to Analog)” found in section 4 of NRSC-G202 “FM IBOC
Total Digital Sideband Power for Various Configurations”.
In this example, the station is operating in Service Mode MP1 with a Nominal D-A power ratio
of -20 dBc. The initial analog reference level of approximately +9 dBm is 22 dB greater than the
digital reference level of approximately -13 dBm. Thus, 2 dB of additional attenuation should
be placed on the analog side of the splitter/combiner so that the digital reference level will be 20
dB below the analog reference level.
Determine whether additional corrections need to be added to adjust for differing coupler ratios,
coupler calibration offsets, and sample line losses. Add (or subtract) the appropriate additional
attenuation to the splitter/combiner device to obtain a correct ratio between analog and digital
signals.
Verify Level Correction
Insert the required attenuation calculated above, reconnect the analog RF sample cable and
disconnect the digital RF sample cable. Restart averaging the then note the analog reference
level. Adjust the reference amplitude so that it is equal to the level indicated by the marker as
shown in Figure 4.
If desired, engage the “delta marker” function with the analog reference level at the top of the
display. This marker can be employed to view the differences in level, in dBc, of any point on
the trace. It avoids having to repeatedly subtract the reading on the trace from the reference level
(both in dBm).
Disconnect the analog RF sample cable and reconnect the digital RF sample cable (leaving the
attenuator pads connected to the splitter/combiner as before), restart averaging and after 50 or
more samples, verify that the digital reference level is the appropriate “dBc” ratio below the
analog reference level as shown in Figure 5.
26
As a quick check, the attenuation required should be equal to the sum of line/filter/combiner losses and antenna
gain (in dB) of the analog transmission system less the sum of the line/filter/combiner losses and antenna gain (in
dB) of the digital transmission system. This assumes that the two directional couplers employed have identical
coupling ratios, that the sample cables have identical loss, and that sampling is done close to the transmitter outputs.
Page 82
NRSC-G201-B
Figure 4
Analog FM Reference Level First Adjusted by Selection of CHIMP Input Attenuation Is
then Set to the Top of the Display
Page 83
NRSC-G201-B
Figure 5
Digital Reference Level Is Observed to Be Certain It Is At the Expected Level
with Respect to the Top of the Display
Page 84
NRSC-G201-B
Perform Measurements
At this point, conventional hybrid FM IBOC measurements can begin by reconnecting the analog
RF sample cable (both analog and digital sample cables should now be connected), change the
Resolution Bandwidth to 1.0 kHz and the Video Bandwidth to 10 kHz or higher. Then restart
the averaging to clear the memory of old display data. After 100 sweeps, the display should look
similar to Figure 6.
Figure 6
Resulting Combined Hybrid FM IBOC Signal Produced by the CHIMP and Measured by
the Analyzer
Page 85
NRSC-G201-B
Diagnostics
In addition to the combined measurements, analog or digital-only measurements can be made by
disconnecting one the RF sample cables. This can be very helpful in determining the cause of
any out-of-compliance conditions. Figure 7 shows an overlay of
the analog only and digital only signals. As can be seen, the “regrowth” products are almost
entirely attributable to the digital transmitter as in this case, the digital transmitter does not utilize
any form of “adaptive pre-correction”. The facilities in this example consist of separate ½-wave
spaced analog and digital antennas which are physically separated by approximately 40 feet. As
a result, the isolation between the antennas is greater than 40 dB.
Figure 7
Analog and Digital Signals Sampled Independently through the CHIMP;
Displayed as Two Traces
Page 86
NRSC-G201-B
Figure 8 shows a more ideal hybrid IBOC display. In this example the facilities consist of
interleaved analog and digital antennas with opposite circular polarization (the analog antenna is
right-hand circular and the digital antenna is left-hand circular). Again, the isolation between the
antennas is greater than 40 dB and in this case, the digital transmitter incorporates “adaptive precorrection”.
Figure 8
CHIMP-based Measurement of a Hybrid FM IBOC Station with Transmission
Characteristics that Differ from the Station Represented by Figures 6 & 7
Page 87
NRSC-G201-B
Figure 9 shows a less-than-ideal hybrid IBOC display. In this example the facilities consist of a
dual-input antenna system where the analog to digital isolation is less than 25 dB. A ferrite
isolator was installed in the digital transmission line to achieve an analog to digital isolation to
approximately 50 dB. Again, in this example, the digital transmitter does not utilize any form of
“adaptive pre-correction”.
Figure 9
CHIMP-based Measurement of a Hybrid FM IBOC Station with Transmission
Characteristics that Differ from the Stations Represented by Figures 6 & 7 and
Conclusion
The Combined Hybrid IBOC Measurement Package (“CHIMP”) as described in this paper is a
reliable tool for evaluating the emissions of hybrid FM IBOC signals that are transmitted by
systems with separate-line architectures. For the CHIMP to provide reliable results, the
following conditions must be met:
•
•
The transmitted power ratio between the analog and the digital components of the hybrid
FM IBOC signal must be properly established with accurate transmitter power, line loss,
and antenna gain information on both the analog and digital transmission chains,
The RF sample from each transmission line must be directional (forward) with sufficient
directivity to prevent measurement error from reverse path crosstalk,
Page 88
NRSC-G201-B
•
•
•
Proper isolation should be designed into the system to prevent measurement error from
reverse path crosstalk (and to prevent unnecessary generation of intermodulation products
within the output of each power amplifier),
The CHIMP should contain suitable pads placed on the inputs to the CHIMP combiner to
provide nearly constant impedance to the output of the CHIMP combiner to enable
reliable measurements when one input has been disconnected.
Variable attenuation at the CHIMP inputs should have sufficient precision to align the
analog and digital power levels according to the calculated transmitted ratio.
Figures 2-7 – WRDA, Canton, GA (Atlanta metro)
Figure 8 – WKXJ, Walden, TN (Chattanooga metro)
Figure 9 – WKSJ-FM, Mobile, AL
Page 89
NRSC-G201-B
ANNEX 5 – RECOMMENDED AM ANTENNA BANDWIDTH CHARACTERISTICS
Page 90
NRSC-G201-B
Recommended AM antenna bandwidth specifications
The recommended specifications in this annex are based on information from Rackley and Dawson in
Reference [14], as presented by Maxson in Chapter 12 of Reference [4].
These specifications represent optimal performance objectives for hybrid AM IBOC antenna system
design. Due to the challenges of antenna system design and implementation, it may not be practicable
for some AM facilities to fully conform with these specifications. Systems not meeting these criteria have
been shown to provide satisfactory hybrid AM IBOC service. The broadcaster should be aware that the
quality and reliability of digital service may diminish as the antenna system performance deviates from
these specifications.
Impedance Bandwidth
Shown in Table 1 are the suggested impedance bandwidth performance specifications for hybrid AM
IBOC stations.
Table 1. Suggested Impedance Bandwidth Performance Specifications for
Hybrid AM IBOC Stations
Frequency range
with respect to reference
channel center frequency
(kHz)
+15 to +10
-15 to -10
−10 to −5
+10 to +5
−5 to +5
Hybrid IBOC digital
signal components
Desired performance
● Primary subcarriers
VSWR below 1.40:1
● Secondary subcarriers
● PIDS subcarriers
VSWR below 1.20:1
● Tertiary subcarriers
● Analog sidebands
● Reference subcarriers
Hermitian Symmetry - impedance at any
frequency within 1.035:1 of the complex
conjugate of the impedance at the mirror
image frequency
The most common method for measuring the impedance bandwidth of an AM antenna system is to
disconnect the transmission line at the transmitter output and attach an impedance measuring instrument
to the antenna system at this point. By attaching at this point, the instrument captures the complex
impedances of transmission lines, phasing and tuning units, and the antenna elements themselves.
However, in this case the instrument does not capture the impedance-changing effects of the
transmitter’s internal components, such as its matching network or its solid state combining network (if
applicable).
It is anticipated that the transmitter manufacturer will supply the broadcaster with a correction figure or
curve to apply to the results of the antenna system impedance bandwidth measurements. This correction
figure or curve should be applied to the measurement result before making a comparison with the
suggested performance specifications in Table 1.
Some contemporary models of AM transmitter incorporate internal impedance bandwidth-measuring
circuitry that relies on true power analysis of the signal. Such devices make impedance bandwidth
analysis more convenient and reliable because they make antenna system measurements without
requiring the transmission line to be disconnected and automatically account for combiner and network
Page 91
NRSC-G201-B
characteristics internal to the transmitter. Because transmitter designs vary, consult the transmitter
manual or the manufacturer for instructions on the best way to obtain a corrected impedance
bandwidth measurement of a station’s antenna system.
Rationale for Impedance Bandwidth Measurement Procedure
The ideal way to examine the antenna system characteristics is to view the transmission path from the
“driving point” of the transmitter, as this is where the antenna system should be optimally matched to the
source impedance of the amplifier's active components. This is desirable because any mismatch
reflects to the amplifier and the consequence of this is distortion. By definition the driving point is situated
at the output of the final amplifier ahead of any matching network.
However, in today’s solid-state transmitters, multiple amplifier modules are combined to produce a higher
output level, such that numerous driving points are combined (with carefully tuned delays) to drive the
output network. This point in the circuit is known as the “amplifier summing point.” The summing point of
multiple amplifiers has a source impedance that is typically not at the target value of 50 ohms (and often
in the single digits). Consequently, the transmitter will often include an internal matching network and/or
RF filtering (typically bandpass or lowpass) intended to transform the combined amplifier output to a
nominal 50 ohm source impedance for driving the antenna system.
By disconnecting the transmission line at the transmitter output and connecting the impedance measuring
device to the antenna system at that point, the measuring device will measure the net impedance of all
antenna system components, excluding that of the (internal to the transmitter) matching network. This
internal matching network introduces a phase rotation which is difficult to characterize since it is
impractical to make an impedance measurement from the amplifier summing point to the matching
network output using an external impedance measuring device, due to the non-standard impedance (at
the matching point) and the difficulty in reliably accessing the matching point with a test connection. This
is why it is necessary for the transmitter manufacturer to supply a correction figure or curve with the
transmitter, to be applied to the results of the antenna system impedance bandwidth measurements.
The final antenna system impedance bandwidth measurements must have the phase rotation of the
internal matching network subtracted from the external measurement to reveal what the amplifier
"sees" at the driving point. It is this phase-corrected impedance measurement that must be evaluated for
bandwidth flatness and Hermitian symmetry. For the most accurate results, it is recommended that the
matching network corrections be presented not simply as a single phase rotation value, but instead as a
curve of phase/amplitude response across the 30 kHz bandwidth of the transmitter centered on the
station’s frequency (or as a complex impedance plot, such as on a Smith Chart).
Hermitian Symmetry
Much has been written to explain Hermitian Symmetry, so this concept is not defined here. In simple
terms, and assuming that the complex impedance-versus-frequency measurement has a curved shape,
the tips of that curve (sometimes called “horns”) would ideally appear on a Smith Chart with the horns
pointing to the left (see Figure 1). In the horns-left condition the load is in the parallel resonant condition
and a voltage source will deliver constant power independent of frequency. In the horns-right condition
the load is in the series resonant condition and a current source will deliver constant power independent
of frequency. Most AM transmitters are voltage sources and should operate better in a horns-left
condition at the driving point.
Page 92
NRSC-G201-B
Figure 1. Example Smith Charts illustrating phase rotation and impedance transformation through an AM transmission system.
Source: Maxson, derived from Figure 12.7 of [4]; based on Rackley, BEC proceedings, 2004 [16].
Page 93
NRSC-G201-B
When adjusting a hybrid AM transmission system for Hermitian symmetry:
1) Determine whether the ideal symmetry for the transmitter is horns-left or horns-right (usually
horns-left);
2) Obtain (from the manufacturer) the nominal phase rotation of the transmitter’s internal combining
and output networks;
3) Measure the impedance bandwidth of the AM transmission system from the transmitter output
point;
4) Adjust the measured curve (obtained in step 3) by the transmitter’s internal phase rotation values
(from step 2) to obtain the impedance bandwidth at the driving point of the power amplifier;
5) Examine the adjusted measurement for its deviation from the specifications in Table 1;
6) Adjust the AM transmission system to obtain adjusted measurements that conform as closely as
practicable to the specifications in Table 1.
Hybrid AM IBOC stations may perform adequately outside the Table 1 performance boundaries, if they do
not deviate too far. The quality of the digital portion of the hybrid AM IBOC signal deteriorates with
increased deviation from the optimal values presented in Table 1.
Broad symmetry across the entire channel is hard to achieve in many antenna system designs. The
recommended symmetry across the ± 5 kHz analog spectrum is most critical because it is this symmetry
that enables the upper and lower sidebands of the tertiary digital signals to self-cancel in the analog
receiver. With good symmetry to ± 5 kHz, analog receivers experience minimal digital interference to the
reception of the analog host.
The recommended ideal Hermitian symmetry value, 1.035:1, is not a VSWR per se, but the difference
between a VSWR at one frequency offset from carrier and the VSWR at the opposite frequency offset
from carrier. For example, if the VSWR at 5 kHz might be 1.2:1 on the upper sideband, this
recommendation says the VSWR on the lower sideband should be between 1.165 and 1.235:1 in one
dimension; in two dimensions it is a circle, whose radius is 1.035:1, around the target 1.2:1 VSWR point
at 5 kHz on a Smith Chart.
It is also helpful to obtain reasonable symmetry on the secondary digital sidebands (5-10 kHz), but it is
less important for minimizing analog reception noise due to the narrow bandpass of common analog
receivers. In addition, good symmetry out to 10 kHz also promotes, in the IBOC receiver, joint decoding
of the upper and lower secondary and tertiary sidebands and self-cancellation of the analog host signal
for optimum digital detection.
Pattern Bandwidth
In addition to the impedance bandwidth issues discussed above, the role of pattern bandwidth in hybrid
AM IBOC system design should be acknowledged. The pattern bandwidth of a directional AM station is
the variation in the station’s radiation pattern depending on the frequency and azimuth. Ideally, the
pattern is identical for every frequency within the station’s necessary bandwidth (±15 kHz for hybrid AM
IBOC operation). In practice, the pattern bandwidth is relatively flat within the main lobe of the pattern,
and may vary substantially at other azimuths. This is the reason that the NRSC-G201 Guideline
recommends all directional hybrid AM IBOC mask measurements be conducted in the main lobe.
NRSC-G201-B
The work of Dawson and Rackley, among others, has resulted in a recommended pattern bandwidth
27
performance for directional AM antenna systems. Dawson and Rackley recommend that, “...throughout
the major coverage areas of the pattern(s), amplitude response is within ± 2 dB and phase response is
within 5 µs (or ±. In simple terms, assuming there is a curve to the impedance bandwidth-by-frequency
measurement 27º) across the entire 30 kHz-wide passband.” This performance is best evaluated by
computer modeling of the antenna system. There is no established practice for performing field
measurement of pattern bandwidth.
27
See, for example “So, What Have We Learned, or What to Do about IBOC,” Stephen S, Lockwood, slides, May 26,
2005, Hatfield and Dawson, www.hatdaw.com.
NRSC-G201-B
ANNEX 6 – AM TRANSMITTER MODULATION TECHNICAL PRIMER FOR NRSC
MEASUREMENT GUIDELINE
NRSC-G201-B
AM Transmitter Modulation Technical Primer
for NRSC Measurement Guideline
Tim Hardy, Head of Engineering, Nautel
Dave Hershberger, Senior Scientist, Continental Electronics
Geoff Mendenhall, VP Transmission Research and Technology, Harris Broadcast Communications
A general, band limited, RF signal may be represented as follows:
s (t ) = i (t ) cos(ω c t ) − q (t ) sin(ω c t ) Eq.1
In Eq.1, i(t) and q(t) constitute the complex envelope in Cartesian form. They are sometimes referred to
as the quadrature components or I and Q for short. For RF signals where the upper and lower sidebands
are symmetrical, such as AM, one of the components may be considered to be zero. It is important to
note that each of the quadrature components generally has bandwidth equal to that of the RF signal, s(t).
There is also a direct linear relationship between the spectra of the baseband complex envelope and the
passband RF such that a linear filter applied to the complex envelope centered at zero frequency will have
the same effect on the passband, centered at the carrier frequency, ωc. For these reasons the quadrature
components are often used to represent the radio signal in a baseband signal processing system.
Another mathematical representation is the polar form:
s (t ) = r (t ) cos(ω c t + ϕ (t )) Eq.2
where:
r (t ) = i 2 (t ) + q 2 (t )
Eq.3
 q (t ) 
 Eq.4
 i (t ) 
ϕ (t ) = tan −1 
In the polar representation, r(t) is the envelope and φ(t) is the phase.
The polar representation is important for AM broadcasting because virtually all modern AM broadcast
transmitters operate using envelope elimination and restoration (EER) also known as the Kahn technique.
All EER transmitters share an important aspect that the audio frequency envelope (<100 kHz) and the
radio frequency phase signal are multiplied together in the final amplifier stage as shown in Eq. 2. Many
transmitters use pulse duration modulation (PDM) to generate the audio frequency envelope and apply it
as a variable power supply voltage (modulator) to the final amplifier stage. Other techniques including
digital synthesis of the RF envelope (example – Harris DX series) may be used to accomplish envelope
modulation but the end result is similar.
The EER approach benefits AM transmitters principally by allowing them to be smaller and operate at
significantly higher efficiencies than the linear amplifiers (classes A, B and A/B) generally used for FM
IBOC and TV broadcast transmitters.
Transmitters using EER may suffer from several types of non-linearities, including AM to AM, AM to
PM, and differential time delay distortion between the envelope and RF phase channels, which increase
NRSC-G201-B
the out-of-band emissions if not carefully corrected for. This is because the two component signals, the
AF envelope and RF phase, have a highly non-linear relationship with the linear I and Q components as
shown by Eq. 3 and Eq. 4. These signals have considerably greater bandwidth than the broadcast signal
s(t) and any misalignment between them when combined in the final amplifier will result in distortion and
increased out-of-band emissions.
To be more specific, any difference of absolute time, non-flat frequency response or non-flat group delay
as well as distortion in either of these signals will result in distortion of the output signal. This is unlike a
linear amplification process where frequency response and group delay errors do not result in distortion.
It is important to note that superposition does not apply in EER amplification because it is a nonlinear
process. The sum of the spectra of the separately transmitted IBOC and AM signal components is not
necessarily the same as the spectrum of the IBOC and AM signal components transmitted together.
Consequently, measurements of occupied bandwidth for just one part of the signal (such as analog only),
while useful, cannot be expected to hold when another part of the total signal is "added."
Many AM IBOC transmitters require a “mag - phase delay” adjustment which corrects for differences in
the overall time delay between the audio frequency envelope and RF phase terms. The difference in time
delay may be introduced, for example, in some PDM-type transmitters by interaction of the modulator
lowpass filter with the varying impedance of different antenna systems. If this is not correctly adjusted,
out-of-band emissions will be increased. However, this adjustment does not correct for differences in
time as a function of frequency (i.e. non-flat group delay). It is also important to note:
•
•
A linear (non-flat) frequency response in the envelope or phase channels will cause distortion
(non-linear) of the final RF term. This will increase the occupied bandwidth and may violate the
spectrum mask. This problem can be attributed to the non-linear relationship between the RF
signal and the envelope and phase signal components;
The envelope and phase bandwidths are much larger than required by analog-only AM
modulation. For IBOC operation, simulations have shown that an envelope bandwidth of 45 to
60 kHz is required. This behavior may also be attributed to the non-linear relationship between
the RF signal and the envelope and phase signal components.
Figure 1 shows an ideal power spectrum of the magnitude term of the hybrid AM IBOC signal (both
digital and analog components present). Note that the envelope signal crosses -80 dBc at 65 kHz. This is
a good indication of required envelope bandwidth.
NRSC-G201-B
Magnitude Spectrum
0
-20
Magnitude (dBc)
-40
-60
-80
-100
-120
0
0.2
0.4
0.6
1.2
1
0.8
Frequency from carrier (Hz)
1.4
1.6
2
1.8
5
x 10
Figure 1. Envelope spectrum of hybrid AM IBOC
AM Audio / IBOC Interaction
It is also useful to consider the time domain behavior of the magnitude and phase signals by referring to
Eq. 3 and Eq. 4. The magnitude term is relatively free of discontinuities in time as it is simply the
magnitude of the sum of the I and Q vectors. However the phase signal is not always so well-behaved.
The phase signal is defined by the arctangent of the quotient of I and Q. When I and Q are both very
small, i.e. the magnitude is near zero, a small change in I or Q may make a large change in phase.
Essentially, the smaller the magnitude of the signal, the higher the rate-of-change of phase, including
nearly instantaneous phase reversals which dramatically increase the short-term bandwidth of the RF
phase signal. Meanwhile, the envelope channel does not have enough modulation bandwidth to
accurately track with the phase channel.
As a result, non linear products resulting in out-of-band emissions from the transmitter tend to be highest
when the magnitude of the signal is small. For this reason, the envelope modulation bandwidth and
linearity of the transmitter is particularly important near the 100% negative modulation point where
carrier pinch-off would occur.
The signal being transmitted is the sum of the IBOC OFDM signal, the AM sidebands and the AM
carrier. The out-of-band emissions are most likely to be large when the analog AM modulation is in the
“trough.” To reduce out-of-band emissions on hybrid AM IBOC transmitters, it is generally
recommended to ensure that the negative analog AM modulation depth is limited to 90% or less. Out-ofband emissions from the transmitter may tend to have bursts associated with these modulation troughs.
NRSC-G201-B
NRSC Document Improvement Proposal
If in the review or use of this document a potential change appears needed for safety, health or technical
reasons, please fill in the appropriate information below and email, mail or fax to:
National Radio Systems Committee
c/o Consumer Technology Association
Technology & Standards Department
1919 S. Eads St.
Arlington, VA 22202
FAX: 703-907-4158
Email: standards@cta.tech
DOCUMENT NO.
DOCUMENT TITLE:
SUBMITTER’S NAME:
TEL:
COMPANY:
FAX:
EMAIL:
ADDRESS:
URGENCY OF CHANGE:
_____ Immediate
_____ At next revision
PROBLEM AREA (ATTACH ADDITIONAL SHEETS IF NECESSARY):
a. Clause Number and/or Drawing:
b. Recommended Changes:
c. Reason/Rationale for Recommendation:
ADDITIONAL REMARKS:
SIGNATURE:
DATE:
FOR NRSC USE ONLY
Date forwarded to NAB S&T:
Responsible Committee:
Co-chairmen:
Date forwarded to co-chairmen:
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising