Radio noise measurements
European harmonisation of measurement methods
This document is produced in cooperation with the following organisations
Agentschap Telecom
NEDAP
Royal Netherlands Navy
VERON
Royal Netherlands Army
ASTRON
Radio Netherlands Worldservice
Date :
14 September 2005
Copyright :
Radiocommunications Agency Netherlands - department of Applied Spectrum Research [TSO] ©2005
Onderdeel Ministerie van Economische Zaken
Agentschap Telecom
Contents List of participants……………………………………………………..….……………. 1 Summary ………………………………………………………….………………….. 2 Introduction …………………………………………………….………..…………… 3 Properties of noise to be measured…………………………….…………………..
4 5 5 6
4 Measurement method / algorithms …………………………..………….….……… 4.1 Receivers and detectors…………………………..………….….……….… 4.2 Selecting a frequency…………………………..………….….……………. 4.3 Collecting the noise containing samples…………………………..………. (Swept or single frequency measurement) 4.4 Selecting the noise containing samples…………………………..…….… 4.5 Correcting for equipment noise…………………………..…………..…….. 4.6 Correcting for 20% or x% values…………………………..………………. 4.7 Correcting for antenna (K) factor differences in scanning.………….….. measurements 4.8 Correcting for filter shape / bandwidth…………………………..………… 4.9 Plotting the results…………………………………………………………… 4.10 Alternative method……………………………………………………………
7 7 9 10
5 Required equipment specifications ……..………………..…………..…….……… 5.1 Receiver specifications……………………………………………………… 5.2 RMS or average detector…………………………………………………… 5.3 Sensitivity…………………………………………………………………….. 5.4 Input impedance…………………………………………………………….. 5.5 Spurious radiation…………………………………………………………… 5.6 Low Noise Amplifier (LNA) and Preselectors…………………………….. 5.7 Cables, cable routers and connectors…………………………………….. 5.8 Feeder identification, terminations and grounding…………………………. 5.9 Sealing………………………………………………………………………… 5.10 Inspection for moisture……………………………………………………..
15 15 16 16 16 16 17 17 18 18 18
6 Antenna systems ……………………………………………..……………………… 6.1 Introduction ……………………………………………..…………………… 6.2 Antenna properties………………………………………..…………………. 6.2.1 Calculation of the antenna factor (K-factor)..……………………. 6.2.2 Gain vs frequency (bandwidth and compensation) ……………. 6.2.3 Measurement sensitivity matters…………………………………. 6.3 Required properties of low noise amplifiers………………………………. 6.4 Methods of calibration of the antenna factor……………………………… 6.4.1 Using a laboratory calibrated measuring antenna……………… 6.4.2 Using simulation tools……………………………………………… 6.5 Field strength over ground or free space fieldstrength………………….. 6.6 Stability (dependency on ground properties) …………………………….. 6.7 Long term stability……………………………………………………………. 6.8 Integrity testing………………………………………………………………. 6.9 Input impedance and measurement error………………………………….. 6.10 Wind loading…………………………………………………………………. 6.11 Wind vibration………………………………………………………………… 6.12 degradation of antenna performance……………………………………… 6.13 Examples……………………………………………………………………. 6.13.1 German quad……………………………………………………… 6.13.2 Inverted V…………………………………………………………... 6.13.3 Monopole…………………………………………………………… 6.13.4 Loop antenna’s………..……………………………………………
19 19 19 19 21 20 20 21 21 21 22 23 23 23 24 24 24 24 24 24 28 30 31
7 Measurement conditions and site survey…………….……………..……..………
32
2
10 13 13 14 14 14 14
Radio noise measurements
7.1 Initial site survey and antenna selection ……….…………………………. 7.2 Calibration and static site survey……………….………………………….
32 33
8 Format for the exchange of measurement data……….………………………….. 8.1 Common interchange format……………………………….………………. 8.1.1 header section……………………………….………………..……. 8.1.2 data section……………………………….………………..………. 8.1.3 example files..……………………………….………………..……. 8.2 Transfer software….……………………………….………………………… 8.3 Presentation of data…………………………………………………………. 8.4 Correlation of multiple 24 hour plots/periods…………………………….. 8.5 365 day plots…………………………..………….….………………..……..
34 34 34 34 34 36 36 37 37
9 Measurement accuracy………………………………….…………….…………….. 9.1 Calibration uncertainty of measurement antenna…….…………….…….. 9.2 Mismatch between receiver and LNA……………………………………… 9.3 Gain accuracy of LNA………………………………………………………… 9.4 Accuracy of external attenuator……………………………………………… 9.5 Calibration uncertainty of receiver / analyser………………………..…….. 9.6 Accuracy of the noise source to determine the correction factors.……… 9.7 Total measurement uncertainty………………………………………………
38 38 38 39 39 40 40 41
ABBREVIATIONS………………………………..……………………….……………..
42
REFERENCES …………………………………..……………………….……………..
43
APPENDIX 1: K-factor calculations…………..……………………….…………….…
44
3
Agentschap Telecom
List of participants The following people have been present in one or more of the meetings of the expert group on noise measurements. In alphabetical order:
Jean Paul van Assche
Radiocommunications Agency Netherlands
Edwin van Bladel
Royal Netherlands Army
Albert Jan Boonstra
ASTRON
Jerry Doms
Koninklijke Landmacht
Koos Fockens
NEDAP / VERON
Chapter 3,4,6,9
Jos Kamer
Radiocommunications Agency Netherlands
Chapter 5,9
Boy Kentrop
Radio Netherlands Worldservice
Henk Klok
Dutch Royal Navy
Chapter 5,7
Erik van Maanen
Radiocommunications Agency Netherlands
Chapter 3,4,6,9,App1
Peter Rozendal
Dutch Royal Navy
Chapter 6,7
Henk Stel
Radiocommunications Agency Netherlands
Chapter 4,8
Henk Vroolijk
DTO (militairy frequency management)
Ben Witvliet
Radiocommunications Agency Netherlands
Ali Hoessein Zadeh
Royal Netherlands Army
Peter Heimenberg
Dutch Royal Navy
Chapter 4,8
App 1
NOTE: Participation of company / institution delegates does not mean that the contents of this document in all cases reflect the company / institution policy about the covered subjects.
4
Radio noise measurements
1 Summary This report describes a new, general, measuring technique for determining the noise floor in practical radio applications. This proposed technique is not equipment-specific; critical elements of the method will be described in detail. To cope with the fact that the method should work with different measurement receivers and analysers, a postprocessing method is developed that works with both scanning and single frequency devices. This method, which will be described in chapter 4, should yield comparable results for both types of devices. Although some examples in the report are based on HF ( 13 dB with the internal attenuator switched off. The output return loss of the antenna is ΓANT > 20 dB. The input and output return loss of a typical LNA is ΓLNA > 7 dB. A schematic presentation of mismatch errors is given in the next picture.
38
Radio noise measurements
Measurement antenna
Low Noise Amplifier
ΓANT
External Measurement attenuator receiver
ΓRX
ΓLNA(22)
ΓLNA(11)
Cable attenuation
Cable attenuation
{
}
The mismatch loss of the receiver to the LNA is A MIS = 10 . log 1 − Γ RX 2 . The mismatch loss of the LNA to the receiver is A MIS = 10 . log 1 − Γ LNA ( 22 )
Mismatch uncertainty is
In absolute terms
U MIS =
U MIS max =
In logarithmic form
2
.
1 1 − ΓRX ΓLNA( 22)
2
1
(1 − ρ RX ρ LNA(22) )2
U MIS = 20 log10 (1 ± ρ RX ρ LNA( 22) )
in dB
The same exercise van be performed for the antenna and LNA input resulting in the following formula
U MIS = 20 log10 (1 ± ρ LNA(11) ρ ANT )
in dB
The attenuation of the cabling is measured and has it’s own calibration uncertainty depending on the calibration equipment used. A way to simplify error analysis is to assume the whole setup: antenna, LNA, external attenuator and cables as a single system. By applying a calibration signal to the input of the LNA the effects of mismatch losses and errors between input LNA and input receiver will be compensated for. This has to be repeated for every measuring frequency (band). This way a single calibration and calibration uncertainty is determined.
9.3 Gain accuracy of LNA The gain of the LNA can easily be determined with a receiver and a generator. Assuming the short term stability of both instruments ideal, the accuracy of the resulting measurement equals the readout accuracy of the receiver which is typical between 0,1 and 0,01 dB. The long term gain stability of the amplifier and the dependency of the gain on the in and output mismatch introduces the dominant error that has to be determined experimentally.
9.4 Accuracy of external attenuator The attenuation of the external attenuator can easily be determined with a receiver and a generator. Assuming the short term stability of both instruments ideal, the accuracy of the resulting measurement equals the readout accuracy of the receiver which is typical between 0,1 and 0,01 dB.
39
Agentschap Telecom
9.5 Calibration uncertainty of receiver / analyser The calibration uncertainty is a standard value given by the manufacturer as u(PM-CAL) typical values are between 0,4 and 1,5 dB (normal distributed, 95% uncertainty). This is a total uncertainty for the whole frequency range and all filter attenuator combinations. The measurement error for this type of measurement can be improved by excluding elements from the receiver error analyses that are not relevant for the measurement. The built in attenuator for example is not used instead an external attenuator is used. Most manufacturers provide information to calculate the new total measurement error/uncertainty. See also reference [18]
9.6 Accuracy of the noise source to determine the correction factors In parts of the calibration of the measurement setup a noise source is used to determine the magnitude of correction factors. Its important that the output power of the noise source is stable. The next picture shows a correctionfactor curve established with an unstable noise source. Noise source 2
1.5
level dBµV/m
1
0.5 Variance is 0.5 dB 0
-0.5
-1 11
11.5
12
12.5
13
13.5
14
14.5
15
unstable noise source, peak variance =0.5 dB, drift > 1dB
The noise source should have gaussian noise properties and the output should be pure noise. The next picture shows the spectrum of a noise source containing carriers at certain frequencies around 5 MHz.
Spectrum of noise source with carrier
40
Radio noise measurements
The last important item is the output impedance of the noise source, a directional coupler with reference load or a network analyser suitable for hot S22 measurements can be used to check this impedance.
9.7 Total measurement uncertainty For the determination of the total measurement uncertainty we need to determine the calibration uncertainty of all elements affecting the measurement accuracy. The following items need to be determined. Uncertainty and distribution Sensitivity coefficient Standard uncertainty of the source Derivation factor Using these info we have to calculate the combined standard uncertainty and the expanded standard uncertainty in %. For a detailed description and a step by step description reference [19], can be used. Also reference [20] and reference [21] contain useful information
41
Agentschap Telecom
Abbreviations dB
deciBel
DSP
Digital Signal Processing
CEPT
European Conference of Postal and Telecommunication Administrations
ECC
European Communications Committee
ERC
European Radiocommunications Committee (part of CEPT)
ETSI
European Telecommunications Standards Institute
FM
Frequency Management
HF
High Frequency
Hz
Hertz
IF
Intermediate Frequency
ITU
International Telecommunication Union
ITU-R
International Telecommunication Union – Radiocommunication
kHz
kilo Hertz
LF
Low Frequency
LNA
Low Noise Amplifier
MF
Medium Frequency
MHz
Mega Hertz
NATO
North Atlantic Treaty Organisation
NVIS
Near Vertical Incident Signal
OATS
Open Area Test Site
PT
Project Team
RMDF
Radio Measurement Data Format
RMS
Root Mean Square
SE35
Spectrum Engineering (workgroup 35 of ERC)
SNR
Signal to Noise Ratio
UTC
Universal Time Coordinate
WGFM
Work Group Frequency Management
WGSE
Work Group Spectrum Engineering
42
Radio noise measurements
References 1
Antennas 2nd edition
Kraus
1988
Fundamentals of Spectrum Analysis using Spectrum Analyser FSA
Rohde & Schwarz
2001
3
Measurement of White Noise Power Density
HP application note 63C
4
Spectrum Analysis (noise measurements)
HP application note 150-4
1974
5
Modern Spectrum Analyser Measurements
Morris Engelson
1991
6
ITU-R SM.331-4 - “Noise and sensitivity of receivers”.
1978
7
ITU-R SM.332-4 - “Selectivity of receivers”.
1978
8
ITU-R P.372-8 - Radio noise.
2003
9
ITU/CCIR-REP 258-5 – Man made noise reports of the CCIR annex to volume X Geneva
1990
2
10
ETSI EG 200 053 – Radio site engineering for radio equipment and systems
11
VERON EMC COMMISSIE Koos Fockens – Calibration of the active antenna PA0KDF
2003
12
RMDF format – A.J. Boonstra Dwingeloo
2000
13
CEPT ERC REPORT 69 - Propagation Model and Interference Range Calculation for Inductive Systems 10 KHz - 30 MHz
1999
14
ITU/CCIR-REP 258-5 - Man made noise, Geneva, Geneva
1990
15
ITU/CCIR-REP 322 - World Distribution and Characteristics of Atmospheric Radio noise
1964
16
ITU-R P-845-3 - HF field-strength measurement
1997
17
ITU-R P-527-3 - Electrical characteristics of the surface of the earth
1992
18
A detailed analysis of EMI test receiver measurement accuracy – Manfred Stecher Rohde & Schwarz
1999
19
UKAS (document M3003)- The expression of uncertainty and confidence in measurement 1997
20
EA-4/02
21
ETS ETR-028 -Radio Equipment and Systems (RES) Uncertainties in the measurement of mobile radio equipment characteristics
1994
22
NIST Technical Note 1297 - Guidelines for Evaluating and Expressing the Uncertainty of NIST Measurement Results
1994
Expression of the uncertainty of measurement in calibration European co-operation for accreditation
1999
43
Agentschap Telecom
Appendix 1: K-FACTOR calculations For many antenna’s the so called K-factor is given. The K-factor is a constant to convert the voltage at the terminals of an antenna (in [dBµV]) to electrical fieldstrength (in [dBµV/m]). This K-factor is useful if we want to use such an antenna for fieldstrength-measurements. The unit of K-factor is [dB(1/m)]. In most cases the gain of an antenna is given in dBi or dBx where x represents the reference antenna. So we have to convert from gain to K-factor. K-factors as used in standardized EMC measurements are a different story. EMC measurement antenna’s are used on a prescribed measurement platform and are also calibrated under these circumstances. The K-factor is not only given in relation to the measurement distance but is also a worst case figure. For antenna-measurements and field strength-measurements these EMC K-factors are of no use. Gain expresses the ratio radiationdensitytestantenna / radiationdensityreferenceantenna in dB. This gain can be determined using the pointing vector S which describes the E as well as the H field being a measure for powerdensity. The pointing vector can be described as follow: S (θ , φ ) = [ E θ2 (θ , φ ) + E φ2 (θ , φ )] / Z 0 , W m-2 . When speaking of powerdensity in a point this S=E2 / Z0 in W m-2. The impedance of free space is Z 0 = ε r . ε 0 . µ r . µ 0 en so for vacuum ε 0 . µ 0 = 376,7Ω . Normally we round this to Z0 = 377 Ω also the number 120π is sometimes used. The K-factor is only about the electrical field in dB. For a conversion we have to transform gain in an absolute number divided by Z0 the root of this number is the gain. However it’s not that simple, the next description can bring some clearance. Imagine the antenna as a transducer that can convert a part of the powerdensity in a 3 dimensional space in power at its terminals. The antenna has its own internal impedance Z0. Next picture clarifies this and shows that free space also has an impedance Z0. Power density Power
P Zo=50Ω
Zo=377Ω
capture area
Ae As told before powerdensity is expressed using the pointing vector S. The total power received by the antenna depends on its capture area Ae
Valid is
AS ⋅ S = P ⇔ 1 / Ae = S / P
⇒ 1 / Ae = ( E 2 / U 2 ) ⋅ (50Ω / 377Ω)
Further And
S = E 2 / 377Ω P = U 2 / 50Ω
⇔ ( E 2 / U 2 ) = (1 / Ae ) ⋅ (377Ω / 50Ω)
⇒K
= E / U = (1 / Ae ) ⋅ (377Ω / 50Ω)
According to Kraus [ref 1]
Ae = (G ⋅ λ 2 ) / 4π ⇔ 1 / Ae = 4π / (G ⋅ λ 2 )
44
⇒ K = 4π / ( g ⋅ λ2 ) ⋅ (377Ω / 50Ω) ⇔ K = 4π ⋅ (377Ω / 50Ω) ⋅ (1 / λ ) ⋅ (1 / G )
Radio noise measurements
⇔
K=
9.73
λ. G
The inverse of this formula can be used to determine a formula to calculate G back to K. ⇔
9.73 G= K. λ
2
This is all valid for a 50 Ω antenna, for other imputimpedances we can derive a set of similar formula’s. Pay attention to the fact that both K-factor and G are presented in linear form and the K-factor is a voltage ratio and gain a power ratio. The next formula is the same as the previous two formulas but converted to dB´s. f= frequency in MHz and G= gain in dBi ⇔
K = 20 log f − G − 29 ,78 dB
45