Nisr UNITED STATESDEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDSAND TECHNOLOGY

NATL INST. OF STAND & TECH R.I.C.

Allia3 M3mMb

NIST

PUBLICATIONSNIST Technical Note 1342

NATIONAL INSTITUTE OF STANDARDS &TECHNOLOGY

Research InformatkMi CenterGaitharsburg, MD 20899

DATE DUE

Demco, Inc. 38-293

Measurement and Evaluation of

a TEM / Reverberating Chamber

Myron L. Crawford

Mark T. MaJohn M. Ladbury

Bill F. Riddle

Electromagnetic Fields Division

Center for Electronics and Electrical Engineering

National Engineering Laboratory

National Institute of Standards and Technology

Boulder, Colorado 80303-3328

Prepared for

Electromagnetic Environmental Effects Division

U.S. Army Electronic Proving GroundFort Huachuca, Arizona 85613-7110

%

U.S. DEPARTMENT OF COMMERCE, Robert A. Mosbacher, Secretary

NATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY, John W. Lyons, Director

IssuedJuly 1990

National Institute of Standards and Technology Technical Note 1342

Natl. Inst. Stand. Technol., Tech. Note 1342, 116 pages (July 1990)

CODEN:NTNOEF

U.S. GOVERNMENT PRINTING OFFICEWASHINGTON: 1990

For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402-9325

CONTENTS

PageList of Figures and Tables v

Abstract 1

1. Introduction and Background 1

2. Anticipated Advantages and Limitations 3

3. Chamber Design and Theoretical Considerations 4

3.1 General Considerations 4

3.2 USAEPG 1/10 Scaled TEM/Reverberatlng Chamber 6

3.3 Proposed Full -Size TEM/Reverberatlng Chamber 7

4. CW Evaluation of the 1/10 Scaled TEM/Reverberatlng Chamber 8

4.1 Measurement Approaches 8

4.1.1 TEM Mode 8

4.1.2 Reverberating Mode 8

4.2 NIST CW Evaluation System - 9

4.3 CW Evaluation and Calibration Results 10

4.3.1 TEM Line and Antenna VSWR and Chamber CouplingEfficiency 10

4.3.2 Chamber Quality Factor 11

4.3.3 E-Fleld Amplitude Calibration 11

4.3.4 Effects of RF Absorber Loading and Multiple DrivenPlates 13

4.3.5 Tuner Effectiveness 144.3.6 Test Zone E- field Uniformity 154.3.7 Test Results from 20 GHz to 40 GHz 16

4.3.8 Summary of CW Calibration and Evaluation Results 17

5. Pulsed RF/Tlme Domain Evaluation of the 1/10 ScaledTEM/Reverberatlng Chamber 18

5 .

1

Background 18

5.2 NIST Pulsed RF Evaluation System 18

5.3 Pulsed RF Evaluation Results 19

5.4 Comments on Pulsed RF Measurement Results 20

6. Projecting the 1/10 Scaled Chamber CW and Pulsed RF MeasurementResults to a Full -Scale TEM/Reverberatlng Chamber 21

7. Performing EMC/V Measurements Using the TEM/Reverberatlng Chamber 22

7.1 Performing EMC/V Measurements Using Manual Operations 22

7 .

2

Automated EMC/V Measurements 25

m

CONTENTS (cont.)

7.2.1 TEM Mode Measurement Procedure 25

7.2.2 Reverberating Mode-Tuned Measurement Procedure 25

7.2.3 Reverberating Mode -Stirred Measurement Procedure 26

7.2.4 Comparison of TEM/Reverberating Chamber MeasurementResults with Anecholc Chamber Tests - Some Examples -- 27

8. Summary and Conclusions 28

9

.

Acknowledgments 30

10. References 30

TV

List of Figures and Tables

Figure 3.1 Cross sectional sketches of the USAEPG 1/10 scaledTEM/reverberating chamber -- - 33

Figure 3.2 Photographs of the USAEPG 1/10 scaledTEM/reverberating chamber -- 34

Figure 3.3 Theoretical mode distribution for the 1/10 scaledTEM/reverberating chamber (1.31 mx 2.41 mx 3.87 m) 35

Figure 3.4 Cross section view of a TEM cell for determiningthe field distribution 36

Figure 3.5 Cross sectional sketches of a proposed full scale TEM/reverberating chamber (13.1 m x 24.1 mx 38.7 m) 37

Figure 3 . 6 Theoretical mode distribution for the proposed full scaleTEM/reverberating chamber (13.1 mx 24.1 mx 38.7 m) 38

Figure 3.7 Theoretical mode distribution for a large TEM/reverberatingchamber (16.2 m x 34.4 m x 38.7 m) 39

Figure 3.8 Theoretical quality factor of proposed full scale TEM/reverberating chamber (13.1 m x 24.1 m x 38.7 m) assuming:(a) galvanized (zinc) steel and (b) enamel painted(cold rolled) steel 40

Table 3.1 Comparison of test fields in a galvanized (zinc finish)steel reverberating chamber of 3.7 mx 5.2 mx 9.8 m withthat in an anechoic chamber at 3 m separation distancefor same transmitted power. Test volume definition inanechoic chamber is (± 3 dB) 41

Table 3.2 Comparison of test fields in a galvanized steelreverberating chamber with that in an anechoic chamberfor same transmitted power. Separation distances inanechoic chamber = 1 and 3 m 42

Table 3.3 Test volume for radiated fields in anechoic chamber at 1 mand 3 m separation distances with selected transmittingantenna gain. Test volume definition is based on half-power points of the radiation pattern 43

Table 3.4 Theoretical mode calculations of frequency and gapbetween adjoining modes for 1/10 scaled TEM/reverberatingchamber, a = 1.31 m, b = 2.41 m, and d = 3.87 m 44

Table 3.5 Normalized electric-field components in the 1/10 scaledTEM cell with 100 Q characteristic impedance, (a) backplate and (b) side plate 45

List of Figures and Tables (cont.)

Table 3.6 Relative permeability of cold rolled 11 gauge steel to beused in construction of welded shielded enclosures as

a function of frequency 46

Figure 4.1 Block diagram of the instrumentation used in the cwevaluation of the USAEPG 1/10 scaled TEM/reverberatingchamber - 47

Figure 4.2 Photograph of the NIST equipment used to evaluate the

USAEPG 1/10 scaled TEM/reverberating chamber 48

Figure 4.3 Measured VSWR of (a) top plate, (b) back plate, and(c) side plate TEM line, from 1 to 2000 MHz, transmittinginto the 1/10 scaled TEM/reverberating chamber 49

Figure 4.4 Measured composite VSWR of the top plate (0.03 to

1.0 GHz) line and the broadband horn antenna (1.0 to

18 GHz) transmitting into the 1/10 scaled TEM/reverberating chamber 50

Figure 4.5 Measured coupling efficiency between the top plate linetransmitting) and log periodic antenna (receiving, 0.1to 1.0 GHz), and the broadband horn antennas (transmittingand receiving, 1.0 to 18 GHz) in the 1/10 scaledTEM/reverberating chamber 51

Figure 4.6 (a) Theoretical and experimental Q of the USAEPG 1/10scaled chamber and (b) ratio of theoretical toexperimental Q, 0.2 to 18 GHz - 52

Figure 4.7 E-field components measured at the center of the 1/10scaled chamber using a NIST isotropic (5 cm dipole)probe (position #3 in figure 4.14) with chamber excitedby: (a) top plate, (b) back plate, and (c) side plate.Net input power is 1 W 53

Figure 4.8 Maximum and average E-field in the 1/10 scaled chambermeasured with (a) calibrated 1 cm dipole probe (0.03 to

18 GHz), and (b) log-periodic receiving antenna (0.1 to

1 GHz) and broadband horn receiving antenna (1 to 18 GHz)

.

Net transmitted power is 1 W - - 54

Figure 4.9 Input power required to generate a 20 V/m maximiom

E-field at the center of the 1/10 scaled chamber using:(a) back plate (Hor. Pol.), (b) top plate (Vert. Pol.),and (c) log periodic and broadband horn 55

Figure 4.10 Cross sectional sketches of the USAEPG 1/10 scaledTEM/reverberating chamber showing placement of 15

pieces 0.2 mx 0.6 mx 0.6 m rf absorber 56

VI

List of Figures and Tables (cont.)

Figure 4.11 E- field components measured at the center of the 1/10scaled chamber using a NIST 5 cm dipole, isotropic,probe with the chamber excited by: (a) top plate, no

absorber; (b) top plate, absorber loaded; (c) backplate, no absorber; (d) back plate, absorber loaded;

(e) side plate, no absorber; (f) side plate, absorberloaded; (g) back and side plates driven simultaneouslyin phase, no absorber; and (h) all three plates drivensimultaneously in phase, no absorber 57

Figure 4.12 E- fields calculated from receiving antenna powermeasurements in the 1/10 scaled chamber with 1 W netinput power. Same transmitting and receivingantennas used as in figure 4.8 59

Figure 4.13 Measured tuner effectiveness in the 1/10 scaledchamber using: (a) receiving antenna power measurements(same receiving antennas as figure 4.5), and (b) 1 cmdipole probe E- field measurements 60

Figure 4.14 Cross sectional sketch of the USAEPG 1/10 scaledTEM/reverberating chamber showing placement of isotropicprobes for measuring spatial E-field uniformity 61

Figure 4.15 Spatial distribution of the E-field components in the

1/10 scaled chamber measured with 7 NIST isotropicprobes (5 cm dipoles) when transmitting 1 W net inputpower from top plate: (a) maximum and (b) average values 62

Figure 4.16 Smoothed average and maximum of measured E-fieldcomponents in the 1/10 scaled chamber generated bythe top plate with 1 W net input power, determinedfrom data in figure 4.15 64

Figure 4.17 Spatial distribution of E-field components in the 1/10scaled chamber measurh 7 NIST isotropic probes(5 cm dipoles) when transmitting 1 W net input powerfrom back plate: (a) maximum and (b) average values 65

Figure 4.18 Smoothed average and maximum of measured E-fieldcomponents in the 1/10 scaled chamber generated bythe back plate with 1 W net input power, determinedfrom data in figure 4.17 67

Figure 4.19 Spatial distribution of E-field components in the 1/10scaled chamber measured with 7 NIST isotropic probes(5 cm dipoles) when transmitting 1 W net input powerfrom side plate: (a) maximum and (b) average values 68

Figure 4.20 Smoothed average and maximum of measured E-fieldcomponents in the 1/10 scaled chamber generated bythe side plate with 1 W net input power, determinedfrom data in figure 4.19 70

vii

List of Figures and Tables (cont.)

Figure 4.21 Measured coupling efficiency (minimum loss) of the hornantennas (transmitting/receiving) placed inside the

1/10 scaled chamber at 20 to 40 GHz 71

Figure 4.22 Maximum E- fields in the 1/10 scaled chamber determinedfrom power measurements of the receiving antenna, 20

to 40 GHz, with a net input power of 1 W 72

Figure 4.23 Calculated normalized propagation attenuation in

air for 5 humidities 73

Table 4.1 Spatial variations in E- fields measured inside the

1/10 scaled TEM/reverberating chamber over test volumedefined by probes shown in figure 4.14 74

Table 4.2 Summary of operation characteristics of 1/10 scaledTEM/reverberating chamber 75

Figure 5.1 Block diagram of instrumentation used in pulsed rfevaluation of the 1/10 scaled TEM/reverberating chamber 76

Figure 5.2 Maximum values of transmitted (positive) and received(negative) rf pulse waveforms in the 1/10 scaled TEM/reverberating chamber determined with increasingamounts of rf absorber loading at . 9 GHz 77

Figure 5.3 Maximum values of transmitted and received rf pulsewaveforms inside empty (no absorber) 1/10 scaled TEM/reverberating chamber determined at selected frequenciesRf input pulse width is 5 /zs , duty cycle is 0.001

Figure 5.4 Maximum values of transmitted and received rf pulsewaveforms inside 1/10 scaled TEM/reverberating chamberdetermined with: (a) no absorber, and (b) 1 piece of20.3 cm X 61 cm X 61 cm (8" x 24" x 24") rf absorberat selected frequencies 79

Figure 5.5 Maximum and average values of received rf pulsewaveforms inside the 1/10 scaled TEM/reverberatingchamber determined by mode -tuned approach. Chamberempty (no absorber) . Measurements were taken at 10

selected frequencies and 3 pulse widths 82

Figure 5.6 Time required for rf signal transmitted into the 1/10scaled TEM/reverberating chamber to rise 63% of steady-state amplitude using: (a) no absorber, and (b) 1 pieceof 20.3 cm X 61 cm X 61 cm rf absorber • 86

vm

List of Figures and Tables (cont.)

Figure 5.7 Time required for rf signal transmitted into the 1/10scaled TEM/reverberating chamber to rise 90% of steady-state amplitude using: (a) no absorber, and (b) 1 pieceof 20.3 cm X 61 cm X 61 cm rf absorber

Figure 5.8 Estimated correction factors for amplitude response of rftest pulses in the 1/10 scaled TEM/reverberating chamberat selected frequencies as a function of input pulsewidth using: (a) no absorber, and (b) 1 piece of20.3 cm X 61 cm X 61 cm rf absorber

Figure 5.9 Estimated correction factors for amplitude response of rftest pulses in the 1/10 scaled TEM/reverberating chamberat selected input pulse widths as a function of frequencyusing (a) no absorber, and (b) 1 piece of 20.3 cmX 61 cm X 61 cm rf absorber 90

Table 5.1

Table 5.2

Figure 6 .

1

Figure 6 .

2

Figure 6 .

3

Table 6.1

Figure 7 .

1

Figure 7.2

Figure 7 .

3

Summary of influences of rf absorber loading on loss,

quality factor, and charge/decay time of the 1/10 scaledTEM/reverberating chamber at . 9 GHz

Comparison of measured and calculated charge/decay timeof the 1/10 scaled TEM/reverberating chamber,determined from chamber average loss measurements(figure 4.5) and from data shown in figure 5.5

Projected coupling efficiency (minimum loss) of 13.1 mX 24.1 m X 38.7 m TEM/reverberating chamber

Projected E- fields in 13.1 m X 24.1 m X 38.7 m TEM/reverberating chamber for 200 W input power determinedfrom (a) projected chamber losses and Q factors, and(b) measurements made in RADC chamber and NSWC chamber

91

92

93

94

Projected charge time for 13.1 m X 24.1 m X 38.7 mTEM/reverberating chamber 95

Summary/cone lus ions - 3 regions of operation -

13.1 m X 24.1 m x 38.7 m TEM/reverberating chamber(zinc coated) - 96

Block diagram of swept frequency EM radiated immunityand shielding effectiveness measurement system usingTEM/reverberating chamber 97

Block diagram of swept frequency EM radiated emissionsand shielding effectiveness system using TEM/reverberating chamber 98

Measured SE data of 3 cm X 6 cm TEM cell with 15 mmcircular aperture using 1/10 scaled TEM/reverberatingchamber and swept frequency measurement system 99

IX

List of Figures and Tables (cont.)

Figure 7.4 Block diagram of automated susceptibility measurementsystem using the TEM/reverberating chamber in the TEMmode of operation 100

Figure 7.5 Block diagram of automated radiated susceptibilitymeasurement system using the TEM/reverberating chamberin the transition and the reverberating modes ofoperation 101

Figure 7.6 Maximum responses of NIST 1 cm dipole probe to anE- field of 37 dBV/m generated in (a) the 1/10 scaledTEM/reverberating chamber, (b) NIST 2.74 m X 3.05 m X4.57 m reverberating chamber, and (c) NIST 4.9 m X6.7 m X 8.5 m anechoic chamber 102

Figure 7.7 Maximum powers received by a broadband (1 to 18 GHz)horn antenna in an E- field of 37 dBV/m using: (a) 1/10scaled TEM/reverberating chamber, and (b) NIST anechoicchamber 103

Figure 7.8 Comparison of broadband horn antenna response in 1/10scaled TEM/reverberating chamber with its responseminus free -space gain measured in the NIST anechoicchamber 104

Figure 7.9 Comparison of SE of 3 cm X 6 cm TEM cell with 15 mmcircular aperture: (a) theoretical calculations,(b) measured in RADC chamber, (c) measured in 1/10scaled chamber with mode -tuned approach, and(d) measured in 1/10 scaled chamber with mode-stirred approach 104

MEASUREMENT AND EVALUATION OF A TEM/REVERBERATING CHAMBER

Myron L. CrawfordMark T. Ma

John M. LadburyBill F. Riddle

Electromagnetic Fields DivisionNational Institute of Standards and Technology

Boulder, CO 80303

This report summarizes the measurement and evaluation of a

1/10 scaled model TEM/reverberating chamber developed as a

single, integrated facility for testing radiated electro-magnetic compatibility / vulnerability (EMC/V) of largesystems over the frequency range, 10 kHz to 40 GHz. Thefacility consists of a large shielded enclosure configuredas a transverse electromagnetic (TEM) , transmission line-driven, reverberating chamber. TEM mode test fields aregenerated at frequencies below multimode cutoff, and mode-stirred test fields are generated at frequencies abovemultimode cutoff. Both the chamber's cw and pulsed rf

characteristics are measured and analyzed. The report alsodiscusses the basis for such a development including thetheoretical concepts, the advantages and limitations, the

experimental approach for evaluating the operational param-eters, and the procedures for using the chamber to performEMC/V measurements. A full-scale chamber that will pro/ide

a test volume of 8 m x 16 m x 30 m is proposed. Some pro-jections of the full-scale chamber's estimated characteris-tics and operational parameters are also given.

Key words: cw and pulsed rf testing; radiated EM com-patability and vulnerability measurements; reverberatingchamber; TEM cell

1. Introduction and Background

The ability to simulate an operating environment for accuratelymeasuring the performance of electronic equipment is fundamental to ensuringthe electromagnetic compatibility (EMC) of the equipment. In recent years,the Fields and Interference Metrology Group of the National Institute ofStandards and Technology (NIST) in Boulder, CO, has been developing improvedmethods and facilities for EMC testing. Two such facilities are thetransverse electromagnetic (TEM) cell and the reverberating (mode-stirred)chamber. Each has its advantages and limitations.

The TEM cell [1,2] is an expanded 50 Q rectangular coaxial transmissionline, with tapered sections at both ends, that operates ideally with onlythe TEM mode to generate the equivalent of a planar field. The cell's useis limited to frequencies below a few hundred megahertz, depending upon thesize of the equipment under test (EUT) and hence the size of the cell. This

limitation is due to the requirement that only the fundamental TEM mode

exist in the cell at frequencies for which tests are to be made.

The reverberating chamber [3,4,5,6] is a shielded room with one or morerotating stirrers to mix the fields transmitted from an antenna placedinside the chamber. The generated test field is determined statistically(maximum and average values) from the multimoded, complex fields. Thistechnique has significant advantages for testing large EUTs efficiently andcost effectively. However, the method is restricted to frequencies above a

few hundred megahertz, again depending upon the size of the test chamber.

This limitation is a consequence of requiring the chamber to be electricallylarge enough that modes exist inside the chamber for adequate mode mixingand spatial averaging to ensure that relatively uniform fields will begenerated for testing.

Combining these two measurement techniques into a single facility willallow EMC testing over the combined frequency range of the TEM cell and the

mode-stirred chamber. This would result in significant cost and timesavings with a possible improvement in measurement accuracy in performingessential EMC tests. A combined TEM/reverberatlng chamber will have threedistinct regions of operation: (a) TEM, (b) transition, and (c)

reverberating. The TEM region will cover the low frequency range near dc to

the first multimode cutoff frequency. The transition region covers therange above the first multimode frequency, up through the frequencies forwhich the first few higher modes are excited In the chamber, butInsufficient in number to provide adequate spatial averaging. The testfield will be a combination of the TEM mode and a limited number of higherorder modes that yield a complex field that is rather difficult to defineboth in polarization and amplitude. Techniques such as absorber loading to

suppress the quality factors of the higher order modes, and/or fieldaveraging over the volume using a matrix of isotropic E- field probes must beused to define the test field parameters. The reverberating region ofoperation will cover the higher frequency range for which sufficient modesexist to ensure adequate mode mixing and spatial averaging. This test fieldsimulates a complex, real -world field environment.

This report summarizes the measurement and evaluation results of a

particular chamber designed to determine how well a combined ("hybrid")TEM/reverberatlng chamber could be used to operate as a radiated EMC/V testfacility. The chamber has the dimension of 1.31 m x 2.41 m x 3.87 m, whichrepresents a 1/10 scaled model of the future full-size facility (13.1 m x24.1 m X 38.7 m). Both the cw and pulsed rf characteristics of this chamberwere evaluated. Anticipated advantages and limitations of this combinedchamber are outlined in Section 2. A description of the chamber Itself is

given in section 3. Section 4 describes the system used for performing thecontinuous wave (cw) evaluation/calibration of the chamber and presents theresults obtained. Section 5 describes the pulsed rf, time -domain evaluationsystems and the associated measurement results. Section 6 provides someprojections, based on the results contained in Sections 4 and 5 for thescaled chamber, of the characteristics for the large, full-scaleTEM/reverberatlng chamber. Section 7 outlines the measurement proceduresfor performing EM susceptibility and shielding effectiveness tests with thischamber by both manual and automated operations . This section also givessample results obtained using control standard EUTs for comparison with data

obtained in an anechoic chamber and other reverberating chambers to showcorrelation factors. Finally, Section 8 gives a sximmary and some

conclusions derived from this study and from previous work in developing the

TEM cell and reverberating chamber measurement techniques

.

2. Anticipated Advantages and Limitations

Motivation for developing a TEM/reverberating chamber arises from its

potential for realizing a number of significant advantages. These include:

a. Electrical isolation from the environment;

b. Accessibility (indoor test facility);

c. The ability to generate high test fields over a large volumeefficiently. For example, 1 W of rf power applied to the

transmitting antenna inside this 1/10 scaled chamber, operating in

the reverberating mode at 10 GHz,generates approximately 100 V/m

maximum test E-field;

d. Broad frequency coverage (10 kHz > 40 GHz)

e. Cost effectiveness. The cost of a TEM/reverberating chamber is

estimated to be less than half that of an equivalent anechoicchamber, primarily due to the cost of rf absorbers and fire-protection system costs associated with an anechoic chamber. Also,rf source power requirements are significantly less for a

TEM/reverberating chamber than for an anechoic chamber as indicatedin section 4. Test time for using the hybrid chamber is anticipatedto be approximately 1/5 of that required when using an anechoicchamber to characterize the maximum response of an EUT

;

f. Reciprocity (the TEM/reverberating chamber can be used for bothimmunity and EMI/TEMPEST emission testing)

;

g. No rotation of the EUT. This results from the three separateorthogonal plates designed for TEM mode operation and the inherenttest field characteristics of the chamber operating in thereverberating mode;

h. Physical security, since the chamber is an enclosed facility; and

i. The potential broad application for cw and pulsed rf testing.

There are also disadvantages or limitations when using such a hybridchamber. Some of these are:

a. Loss of directional or polarization characteristics of the EUTplaced inside the chamber when operating in the reverberating mode;

b. Limitations of the test field to planar fields at lower frequencies(TEM mode of operation) , and to complex multimode fields at higherfrequencies (reverberating mode of operation)

;

c

.

Measurement of only the total power radiated from the EUT in the

reverberating mode of operation as opposed to determining both the

total power radiated and the directional characteristics of the

emissions for TEM mode of operation;

d. Less response of an EUT measured in the reverberating mode than whenmeasured inside an anechoic chamber (free -space) , if the EUT is

exposed to the same level of test field in each chamber. Thisreduction in response is proportional to the EUT's free -spacedirectivity [5]. Hence, the susceptibility criteria determined foran EUT measured in a TEM/reverberating chamber operating in thereverberating mode must include an additional factor proportional to

the EUT's estimated free-space maximixm gain, if the results are to

be correlated to the worst case free -space conditions; and

e. Relatively large measurement uncertainty (as great as ± 10 dB) forestablishing the test field in the frequency range of the transitionregion between TEM mode of operation and reverberating mode ofoperation of the chamber.

3. Chamber Design and Theoretical Considerations

3.1 General Considerations

Factors which influence the chamber's electrical and operationalcharacteristics include: (a) the size and materials used in the chamber'sconstruction, (b) the characteristic impedance of the transmission linesused for TEM mode operation of the chamber, and (c) EUT's mechanical andelectrical operational requirements

.

The size, shape, and materials used in the chamber determine the testvolume, quality factor, and the shielding effectiveness characteristics ofthe chamber. The size and shape also influence the frequency distribution,number of modes, and density of the modes that are excited inside thechamber. At low frequencies, only the fundamental or TEM mode can exist inthe chamber. As the frequency increases, higher order modes may begin topropagate. The resonant frequencies of these modes can be determined by

where i, n, and p are integers, and f. is in megahertz when the

rectangular chamber's internal dimensions a, b, and d are expressed inmeters. By carefully selecting the chamber dimensions, we can minimizeredundant modes (modes occurring at the same frequency) , resulting in abetter mode distribution.

The quality factor of the resonant modes in the chamber determines theinput power requirements for establishing test fields when the chamber isoperated in the reverberating mode. The theoretical composite qualityfactor of a reverberating chamber is approximately [7, 8],

Q = J_ J_ 1(2)

"^ 1 -^16 ^ a

-^b

-^a )

where V is the chamber's volume in m^ , S is the internal surface area in m^

,

5 = J 2/ {(x>iJ.a) is the skin depth of the chamber wall in m, cr is the wall

conductivity in S/m, /i =fj, /j. is the wall permeability in H/m, /j, = Un(lO)

H/m is the free -space permeability, jj, is the relative permeability, and A

is the wavelength in m. This composite Q is determined by averaging the Qvalues of all possible modes within a small frequency interval around the

frequency of interest [7]. The composite Q estimated from (2) is consideredan upper bound because it does not take into account losses other than thatdue to wall conductivity. In reality, other losses will also occur duringthe measurement, for example, due to leakage from the chamber, antennasupport structures, and the chamber's wall coatings. An alternative meansof determining the chamber's Q is also available by measuring the chamber'sloss [9]. Chamber loss is determined experimentally by measuring the

difference between the net input power, P , delivered to the chamber's

transmitting antenna or transmission lines and the power available, P , at a

reference antenna or an EUT. Both P and P are measured in the same unitst r

(W or mW) .

If the energy were uniformly distributed over the volume of the

chamber, an empirical value of the chamber's quality factor (Q') could beobtained using the following equation [9],

V PQ'= 16 7r2 -^ -^

, (3)

A t

where V and A are as previously defined.

If a chamber factor, K = Q/Q' , is determined, the "equivalent" powerdensity, Pj' , in a reverberating chamber may be determined by: [5]

,

47rP

P^' = -^ W/m2 . (4)

With the aid of (3), we can relate P ' to the net input power, P , by

3AP

s '^r

As can be seen from (5) , the material used in constructing the chamber aswell as the chamber's dimensions influence the input power requirements forestablishing a specified test field in the chamber. Typical K factors forreverberating chambers are in the range of 2 to 5.

Obviously, the electrical properties of the materials also influence

the shielding characteristics of the chamber. Since the chamber is

anticipated for use at frequencies as low as 10 kHz, steel with a relatively-

high permeability may be required to obtain adequate shielding from magnetic

fields. If such metal is used, it may be advisable to treat it with zinc

(i.e., galvanize it) or line the chamber with a metal such as aluminum or

copper to improve the surface conductivity and hence the chamber's qualityfactor.

It is of interest to compare the power densities of test fields

anticipated inside a reverberating chamber with that anticipated inside ananechoic chamber assuming the same available input power. The powerdensity, P., generated inside an anechoic chamber, assuming far-field

conditions, can be determined by

P G

P^ = ^2 W/m2, (6)

dk'K r

where G is the free -space, far -field gain of the transmitting antenna and r

is the separation distance between the antenna and the test location in m.

A comparison of the test fields given as the ratio, '?'/'?, obtained from

(5) and (6) for a comparable size, zinc -finish, reverberating chamber of 3.7

m X 5.2 m x 9.8 m and an anechoic chamber is given in Table 3.1 at

frequencies from 100 MHz to 500 MHz when the tuned dipoles are used. Thetable also gives the size of the test volume that falls within ± 3 dB regionin the anechoic chamber at r = 3 m. The ratios, P//P., for r = 1 m and 3 m

d dobtained using open-ended waveguide and horn transmitting antennas in ananechoic chamber, at frequencies from 0.2 to 18 GHz, are given in Table 3.2.

The test volume definitions for the radiated fields in the anechoic chamberas a function of antenna gain at 1 m and 3 m separation distances are givenin Table 3.3. These tables demonstrate that significantly less power is

required to generate the same maximum amplitude test fields in a

reverberating chamber over large volumes than it is generated in an anechoicchamber over comparable volumes

.

3.2 USAEPG 1/10 Scaled TEM/Reverberating Chamber

The 1/10 scaled chamber is the model of a proposed largeTEM/reverberating chamber anticipated to be built at the USAEPG, Ft.

Huachuca, AZ . It is a rectangular shielded enclosure, 1.31 m x 2.41 m x3.87 m (4.3 ft x 7.9 ft x 12.7 ft) in size, constructed from sheet steelwelded together at the edges. The interior finish is of flame -sprayed zinc.Cross -sectional drawings of the 1/10 scaled chamber are shown in figure 3.1.Photographs showing different views are shown in figure 3.2. Calculatedresonant-mode frequencies for this scaled chamber using (1) are given intable 3.4. The theoretical mode distribution obtained from these data isshown in figure 3.3 in terms of the percentage frequency interval (gap)between adjoining modes as a function of frequency.

Three orthogonal plates are mounted in the chamber, each designed tooperate as a 100 f2 characteristic impedance TEM transmission line terminated

in its characteristic impedance. One of the cross -sectional views along

with the necessary equations [10] used for determining the electric -fielddistributions inside the chamber are given in figure 3.4.

The test volume in the chamber is the region centered beneath the top

plate and the floor, and between the back and side plates and theircorresponding front and end walls. The size of the usable test volume is

determined by the mode of operation of the chamber. For TEM mode of

operation (approximately from 10 kHz to 70 MHz), the maximum recommendedtest volume is from 1/2 to 1/3 the region between the plates and oppositewalls. For reverberating mode of operation (from 300 MHz up), the testvolume is the total region between the plates and opposite walls minusapproximately 30 cm separation distance from the plates and walls (or 1/6wavelength at frequencies above 300 MHz) . For reverberating mode ofoperation, this corresponds to approximately 0.7mxl.6mx3.0m for the

1/10 scaled chamber.

The theoretically estimated electric- field amplitude distributioninside the chamber, operating in the TEM mode, is given in table 3.5 for the

back plate and side plates driven as TEM transmission lines. The electric-field amplitudes are given in terms of the orthogonal components, defined as

X (tangential to the plate) and y (normal to the plate) relative to theplate of interest. The distribution inside each cross section is relativeto the center value at x = and z = h./2 (back plate) and y = and x =

h./2 (side plate). The theory used to obtain these estimates [10] is not

valid for determining the field distribution for the top plate driven as a

transmission line due to the large gap distance between the plate's sides to

the chamber end walls. This field distribution was determinedexperimentally, as described in Section 4.

Two tuners are used for stirring or redistributing the reflected energyassociated with modes excited inside the chamber for the reverberating modeof operation. The tuners consist of 4 rectangular blades, 0.2 m x 0.4 m incross section, made from plastic foam covered with heavy (> 25 mm) aluminum

foil. The blades are mounted at offset angles (30 to 45 ) with respect to

the steel drive shafts that are connected to precision stepping motors.Each tuner motor operates independently, under computer control, so that thetuner revolution rates and positions can be controlled individually or incombination.

The chamber is equipped with a bulkhead access panel located on theback and a 0.762 m high x 2.286 m wide access door located on the front sideof the chamber.

3.3 Proposed Full-Size TEM/Reverberating Chamber

The proposed full-scale TEM/reverberating chamber will have the overalldimensions of 13.1 m x 24 . 1 m x 38 . 7 m as shown in figure 3.5. It has anapproximate test volume of 8 m x 16 m x 30 m. There are two orthogonalplates (top and back) , each designed to operate with 75 Q characteristicImpedance for the TEM mode of operation. The third (side) plate is deletedbecause of the large gradients expected in the field distribution as

evidenced in the scaled chamber results that were considered excessive. The

reason for using 75 Q instead of 100 Q for the characteristic impedance is

to improve the uniformity of the TEM field distribution in the test volumeas compared to what was realized in the 1/10 scaled model. This will,

however, reduce the field strength by an amount proportional to the squareroot of the ratio of the impedances /(100/75) for the same input power to the

transmission lines. As with the scaled model, the plates will be terminatedin their characteristic impedances and driven independently from a high-power rf source when operated as TEM transmission lines. For the

reverberating mode of operation, the percent gap between predicted modes for

this proposed full-scale chamber is given in figure 3.6. Figure 3.7 givesthe percent gap that would exist if the dimensions of the chamber wereincreased to 16.2 m x 34.4 m x 38.7 m. This larger chamber could be used,

in the reverberating mode, at frequencies down to approximately 20 MHz as

opposed to 30 MHz estimated for the chamber proposed in figure 3.6. The

theoretical composite Q for the full-scale chamber is estimated in figure3.8 for both cold rolled steel and zinc finishes. The values of u used in

r

calculating the theoretical composite Q for a steel chamber are given intable 3.6. The value of fi for a galvanized steel (zinc coated) chamber is

assumed to be 1 . The conductivities (a) used for steel and zinc are 0.4 x

10 S/m and 1.74 x 10 S/m, respectively. The zinc increases the Q byapproximately the square root of the ratio of the conductivities of the twometals at frequencies where

fj,is 1.

4. CW Evaluation of the 1/10 Scaled TEM/Reverberating Chamber

4.1 Measurement Approaches

4.1.1 TEM Mode

For TEM mode of operation in the frequency range from 10 kHz to 70 MHz,two techniques are employed in determining and monitoring the E- field beinggenerated in the chamber. The first technique uses measurements of theincident and reflected powers on the sidearms of a bidirectional coupler,the impedance of the TEM transmission lines, the separation distance betweenthe plates and ground plane, and the location between the plates(anticipated placement of the EUT) to determine the E- field normal to thedriven plate. The second technique is to place a calibrated isotropic E-

field probe in the area where the field is to be measured. Using these twotechniques gives a double check to help assure consistency and validity ofmeasured data while providing a real-time display of the field strength asmeasurements are being made.

4.1.2 Reverberating Mode

For reverberating mode of operation at 300 MHz or higher frequencies,the chamber was evaluated using two different operational approachesreferred to as mode -tuned and mode -stirred [5].

For the mode- tuned tests, the tuner is stepped at selected, uniformincrements. The net input power supplied to the transmitting antenna, the

8

power received by a reference antenna, the field-measuring probe responses

and the EUT response at each tuner position are measured and recorded.

Corrections are then made for the changes in the transmitting antenna's

input impedance as a function of tuner position and frequency. The

measurement results are then normalized to a constant net input power value.

The number of tuner steps used per revolution was 200 (increments of 1.8 ).

For the mode-stirred tests, the tuner was continuously rotated whilesampling the power received by the reference antenna, the field proberesponse, and the EUT response at rates much faster than the tunerrevolution rate. These measurements were made using a spectrum analyzer and"smart" voltmeters with diode detectors. The "smart" voltmeters are capableof data storage and calculation of various arithmetic functions. Large datasamples (up to 9999) are obtained for a single tuner revolution. Tunerrevolution rates are adjusted to meet the EUT output monitor, diode proberesponse time, and sampling- rate requirements of the instrumentation.Typical rates used were approximately 2 to 4 minutes per revolution. Theinput power was measured only at the beginning of each measurement cycle.For the cw tests, a calibrated bidirectional coupler was used to measure thenet input power to the transmitting antenna.

The field level inside the chamber is determined using two techniquessimilar to TEM mode of operation. One technique uses a reference antenna tomeasure the power density of the field inside the chamber by use of (4)

given in section 3.1. The other technique uses one or more calibratedprobes to measure the E- field. These measurements are described in moredetail later.

4.2 NIST CW Evaluation System

The block diagram of the system used in the cw evaluation is shown infigure 4.1. A photograph of the NIST equipment used to perform the chamberevaluations is shown in figure 4.2. The test field is established insidethe chamber by means of an rf source connected to the appropriatetransmitting antenna or plate. Placement of an EUT should fall within thetest volume as defined in section 3.2 except, possibly, relative to thefloor. The separation distance between the EUT and the floor may be lessthan 30 cm depending upon the intended EUT configuration rel^ive to theground plane. (For some applications, the EUT is mounted close to ordirectly on a ground plane. In these cases the EUT should be mounted thesame in the chamber for test purposes.) Test leads and cables are routed to

appropriate monitors, for example, outside the chamber through shieldedfeed- through connectors. Coaxial cables are used for rf signals, and high-resistance lines are used for dc signals. This is done to prevent leakage ofthe EM energy to the outside environment or into the instrumentation room.A 10 -dB attenuator is used whenever possible between the receiving antennaand the 50 Q power sensor or spectrum analyzer. This is done to minimizeimpedance mismatch with the receiving antenna and measurement systemoverload.

The transmitting and reference receiving antennas used for the cw testsare the top, back and side plate transmission lines (1.0 MHz to 1 GHztransmitting), a log periodic antenna (0.2 to 1.0 GHz receiving), and

broadband horn antennas (1.0 to 18.0 GHz transmitting and receiving). Areference receiving antenna is not used at frequencies below 200 MHz. For

these frequencies the field was determined only using calibrated probes,

4.3 CW Evaluation and Calibration Results

4.3.1 TEM Line and Antenna VSWR and Chamber Coupling Efficiency

The efficiency with which energy can be injected into or coupled out ofthe chamber by the transmitting transmission lines or antennas and the

receiving antennas is determined by: (a) the VSWR of the lines or antennas[that is, the impedance match between the rf source and the transmittingline or antenna (load) or between the receiving antenna (acting as the

source) and its output detector (load)], and (b) the ability of the lines or

antennas to couple energy into or out of the particular modes (TEM or

higher order modes) that exist at the test frequencies of interest. For the

reverberating mode of operation, rotating the tuner changes the

characteristics of the field inside the chamber, which in turn influencesthe effective in-situ VSWR of the antennas. Hence the net input power to

the chamber and the received power detected by the reference antenna vary as

a function of tuner position. That is, the impedance matches between the rfsource and transmitting antenna, and between the receiving antenna and its

termination, affect the power transfer between the two antennas. This canresult in errors in determining the amplitude of the field inside thechamber if no correction is made. The equation for calculating themagnitude of the error caused by the impedance mismatch is

2 2fraction of maximum (1 - |r| ) (1 - |r| )

P„ = available power absorbed = ^r, (7)

by the load11 - T T \

where F and F denote the complex reflection coefficients for the sources

and j(p.oads respectively as defined above. Their magnitudes, |F|and |F |,

can be obtained from the appropriate VSWR by the expression,

F. = (VSWR. - 1)/(VSWR. + 1), i = S or L.

The VSWRs of the transmission lines (plates) and the broadband ridged hornantennas used to excite the chamber are shown in figures 4.3 and 4.4.Figure 4.3 gives the VSWRs of the top, back and side plates; and figure 4.4is the composite VSWR of the top plate (1.0 MHz - 1.0 GHz) and the broadbandridged horn antenna (1.0 - 18 GHz) in the transmission mode. The receivingantenna's (log periodic, 0.2 - 1.0 GHz, and broadband ridged horn, 1.0 - 18GHz) VSWRs are similar to figure 4.4. The figures show the maximum,average , and minimum VSWR obtained by rotating the tuner through a completerevolution (360°). Large variations and high values of VSWR exist,especially at the lower frequencies. At higher frequencies (> 2.0 GHz) theVSWRs' variations become less and their values are lower, approaching thefree-space VSWR of the antennas. At frequencies below approximately 200MHz, the VSWR of the lines is unaffected by the tuner rotation. At

10

frequencies above approximately 10 MHz the high VSWR results from modingeffects in the chamber and line impedance mismatch.

The coupling efficiency of the chamber is defined as the ratio of the

net input power delivered to the transmitting plate or antenna to the poweravailable at a 50 fi matched terminals of the receiving antenna. Theseratios, called the chamber loss, are given in figure 4.5. The curves showthe average and minimiom losses measured by the mode -tuned approach, usingthe top plate (0.1 - 1.0 GHz) and broadband ridged horn (1.0 - 18 GHz)

antennas transmitting, with a log periodic (0.1 - 1.0 GHz) and broadbandridged horn (1.0 - 18 GHz) antennas receiving. Impedance mismatch betweenthe receiving antenna and power detector used to measure the received poweris not accounted for in these measurements . The magnitude of the error(uncertainty) resulting from this is discussed in [2, 5, 11] and is includedin table 4.2.

4.3.2 Chamber Quality Factor

The chamber's quality factor, Q, Influences the operation of the

chamber in a number of ways for the reverberating region of operation.Examples are tuner effectiveness, rf input power requirements and the

accuracy with which test field levels can be established inside the chamber.Chamber Q also influences the time required for the chamber to charge up to

a steady-state condition after the input signal is applied. This affectsthe chamber's response characteristics for pulsed rf testing, as is

discussed later.

Results obtained by using (2) to calculate the composite Q with therelative permeability values given in table 3.6, and by using the chamberminimum loss data shown in figure 4.5 together with (3) to determine theexperimental Q' are shown in figure 4.6. Figure 4.6 (a) gives thetheoretical and experimental Q for the chamber, and figure 4.6 (b) gives theratio of theoretical to experimental Q. The * and + points indicate thedata determined from calculations and experimental measurements , and thesolid lines are estimated curve fits. At the higher frequencies the ratioapproaches roughly 2.4. This indicates that at frequencies where (2) and(3) are valid, there are relatively small losses other than the loss in thechamber walls and in the transmitting and receiving plates or antennas(antenna efficiencies)

.

The value of 2.4 is determined from using the conductivity of steel,instead of zinc, in (2) as expected for the zinc flame -spray finish in thechamber. Apparently, the flame -spray finish is too rough compared to theskin depth thickness, giving either the effect of a greatly increasedsurface area or increased loss, thus making the effective finish equivalentto cold rolled steel. Apparently, this type of finish will not give thedesired improvement in Q as anticipated for the higher conductivity zincover steel.

4.3.3 E-Field Amplitude Calibration

The E- field strength in the chamber can be determined either frommeasurements of the input power to the transmission lines (TEM operation) or

n

from the power received at the terminal of the receiving antennas

(reverberating operation) . It can also be determined by measuring the E-

field with calibrated probes.

For low frequencies (TEM region of operation) , the electric field canbe calculated by using the expression,

J P.R1

E = ^ , V/m (8)

where b is the separation distance between the plate and the opposite wallor floor in m, P. is the net input power to the line in W, and R is the

transmission line's impedance (~ 100 Q)

.

For high frequencies (reverberating region of operation) , an equivalentelectric field in the chamber may be calculated using the expression,

^a = J~^ " ^ J ^° ^r '

^/"^ ^^^

where 77 is the average wave impedance of the chamber (377 Q) , P ,' is the

equivalent power density in W/m^ , in the enclosure, A is the wavelength in

m, and P is the measured average received power in W. The validity of (9)

has been verified and discussed in [5]

.

Results of the measured electric fields in the chamber for 1 W netinput power are shown in figures 4.7 and 4.8. Figure 4.7 gives the averageand maximum electric field components, obtained with a calibrated, 3-,

orthogonal -dipole probe placed at the center of the test zone, with thechamber excited by one of the three transmission line plates. The x, y, andz components are the horizontal longitudinal (end to end) , the vertical (topto bottom) , and the horizontal transverse (front to back) , respectively. Inaddition, the total magnitude (RSS value of the three components) is alsoincluded in figure 4.7 as marked by T. Figure 4.8 gives the maximum andaverage electric fields determined from received power measurements togetherwith (9) and from measurements that used a calibrated 1 cm dipole probe.For these measurements the chamber was excited using the top plate atfrequencies from 0.03 to 1.0 GHz and using a broadband ridged horn antennaat frequencies from 1.0 GHz to 18 GHz. For both figures, the tuners were

rotated through one complete cycle (360 )

.

Three distinct regions of operation are noted for this scaled chamber.At frequencies below approximately 70 MHz the field is singly moded (TEMmode) and the tuner rotation has little or no effect upon the field'samplitude. Hence, the maximum and average E- field are the same. Also, thedominant component, the one that is polarization-matched with the TEM modeE- field, is evident. The amplitude of the component corresponds to thecalculated value determined from measuring the net input power and (8). Atfrequencies between approximately 70 MHz and 300 MHz, the chamber undergoes

12

a transition region where higher order modes exist in addition to the TEMmode but in insufficient numbers for the chamber to reverberate properly.

Large variations in the field strength exist in this region. At frequenciesabove approximately 300 MHz, sufficient modes are generated so that the

chamber is reverberating effectively. In this region the effect of Q on the

resonant modes is evident and the tuner is effective in redistributing the

fields within the chamber. Thus the maximum amplitude of the field is

substantially higher than that for the TEM mode of operation (as great as 25

dB) for the same net input power of 1 W, and the polarization properties of

the components are lost. Note that the amplitudes of the three componentsin figure 4.7 in this frequency region are approximately the same, and that

the total magnitude (RSS value) is about 4.8 dB higher than each individualcomponent.

The 3 orthogonal dipole probes and the 1 cm single dipole probe used to

measure the fields in figures 4.7 and 4.8 were calibrated in a planar fieldusing a TEM cell [12] at frequencies up to 500 MHz and in an anechoicchamber at frequencies from 500 MHz to 18 GHz [13]. The assumption is madethat the average field over the aperture of the probe inside the

reverberating chamber will approximate the planar field used to calibratethe probe. This is reasonable, at least at frequencies for which the probeis electrically small. Also, the open-space far-field gain of anelectrically small dipole is small (1.76 dB) . Thus, the probe-measuredfields should be equivalent, within approximately 1.76 dB, to the E- fieldsdetermined using a receiving antenna. This is true if the polarizationproperties of the probe and receiving antenna are effectively eliminated inthe chamber. The variations in the fields determined by using receivedpower measurements and by using the probes (figure 4.8) are typical of therandom variations in the data used to determine the field strength insidereverberating chambers. The agreement between the two methods is consideredreasonable

.

Using the results of figures 4.7 and 4.8, we can calculate the inputpower requirements for establishing a given field level in the chamber. Anexample is shown in figure 4.9, which gives the net input power required togenerate a 20 V/m field in the 1/10 scaled chamber using the top and backplates and a combination of log periodic and ridged horn antennas.

4.3.4 Effects of RF Absorber Loading and Multiple Driven Plates

Tests were made to determine the effect of lowering the Q in thechamber by adding rf absorbing materials, or of driving more than one platesimultaneously, upon the field in the test zone. The transition region ofoperation is of particular interest.

Fifteen pieces of 0.2 m x0.6mx0.6m pyramidal rf absorber wereplaced inside the chamber as shown in figure 4.10. Measurements were thenmade using a 5 cm dipole, isotropic probe placed at the center of the testzone similar to the measurement for figure 4.7 with each of the platesdriven with 1 W net input power. The results are shown as (b)

,(d) , and (f)

of figure 4.11 for the frequency range, 30 to 300 MHz (includes the entiretransition region of operation). Results of driving all the plates and justthe back and side plates, simultaneously from the same source, withoutabsorber loading, are also shown in figure 4.11 as (h) and (g) . These are

13

compared to the results obtained in figure A , 7 , which are reproduced in

figure 4.11 for the 30 to 300 MHz region, as (a), (c) and (e)

.

A comparison of (a) with (b) ,(c) with (d) , and (e) with (f ) shows the

effect of the absorber loading, lowering both the amplitude and the

variation in the amplitude of the measured E-field components. The

absorber's effect is most pronounced at the higher frequencies where the

absorber is most effective in lowering the Q of higher order moderesonances. Thus, use of absorber would be helpful in reducing the

variation in the E-field and hence the uncertainty in determining the

amplitude of the test field. However, this must be balanced against the

reduction in the field level and the cost and inconvenience of placing the

absorber in the chamber. At frequencies above 300 MHz, the absorber wouldhave to be removed from the chamber for proper operation of the chamber in

the reverberating region.

Comparisons of figure 4.11 (a) [top plate, no absorber] with (h) [all

plates, no absorber] and (c) [back plate, no absorber] with (g) [back andside plates, no absorber] indicate similar field levels and variations in

the components. Apparently, little is gained by multiple excitation of the

plates

.

4.3.5 Tuner Effectiveness

Proper operation of the chamber in the reverberating mode is dependentupon the effectiveness of the tuner to obtain randomness in the distributionof the test field inside the chamber. To achieve this the tuners must beelectrically large and shaped or oriented to distribute energy equally intoall chamber resonant modes as much as possible. A test to determine howwell the tuners are functioning is to measure the ratio of the maximum to

minimum E-field in the test volume in the chamber as a function of tunerposition. This is done by either measuring the ratio of the received powerat the terminals of the receiving antenna or by measuring the electric fieldusing a calibrated probe, relative to a constant net input power to thetransmitting antenna. A large ratio (greater than 20 dB) indicates that thetuner is effective in redistributing the scattered fields inside thechamber

.

Measurement results for the scaled chamber are given in figures 4.12and 4.13. Figure 4.12 gives the maximum, average, and minimum E-fieldbased on the receiving antenna power measurements for a complete revolutionof the tuners as a function of frequency. Figure 4.13(a) is the ratio ofthe maximum to minimum E-field, or the tuner effectiveness, determined fromfigure 4.12, and figure 4.13(b) is the corresponding tuner effectivenessdetermined from E-field measurements made using a calibrated 1 cm dipoleprobe. The measurements were obtained by using the same transmitting andreceiving antennas as noted in figure 4.5. This ratio is also influenced,in addition to the tuner size and shape, by the chamber Q. For example,placing rf absorber and/or an EUT inside the chamber will lower thechamber's Q, referred to as the loading effect. A minimum ratio of 20 dB is

recommended to assure proper operation of the chamber [5]. As seen infigure 4.13, a tuner effectiveness of approximately 30 dB is achieved atfrequencies above about 300 MHz for the 1/10 scaled chamber.

14

4.3.6 Test Zone E-Field Uniformity

Tests were made to determine the E- field uniformity in the chamber as a

function of spatial position and frequency. Seven small NIST isotropicprobes [14, 15, 16], designed to operate at frequencies up to 2 GHz, wereplaced at various locations inside the chamber (positions 1, 2, 3, 4, 5, 6,

8; no position 7 was designated) as shown in figure 4.14. Each probe hasthree orthogonally oriented dipoles which are aligned with the three axes of

the chamber, earlier called the horizontal longitudinal (E ), vertical (E ),

and horizontal transverse (E ) . Measurements were made of the fieldz

strength of each orthogonal component at the seven locations for each tunerposition (200 steps of 1.8°) for frequencies from 1 to 2000 MHz. These datawere normalized for a net input power of 1 W applied at the input terminalsof the transmission lines or transmitting antennas. The maximum and averagevalues for each component and the RSS value were then determined from the

complete data sets . The results of these measurements obtained when drivingthe chamber with the top, back, and side plates are shown in figures 4.15thru 4.20. Figures 4.15, 4.17, and 4.19 give the individual data obtainedfor all seven probes. Figures 4.16, 4.18 and 4.20 give the averaged valuesdetermined from the 7 individual data sets of figures 4.15, 4.17, and 4.19.

Again, the three regions of operation are apparent. The spread in the datashows spatial field variations inside the chamber at the indicatedfrequencies. Note that the dominant components match with the TEM mode E-

field polarization for each driven plate in the TEM mode of frequency region(below 70 MHz). For example, the E component is dominant for the top

plate; E is dominant for the back plate; and E is dominant for the side

plate. These results are in agreement with figure 4.7, as expected.Gradients in the E- fields between the plates are significant. This is dueto the placement of the plates close to the top, back, and side walls asnecessary for a clear test volume. The vertical E- field difference for thetop plate, figure 4.15, over 1/2 the vertical separation distance (positions2 to 3 and 3 to 4) , is approximately 12 dB (± 6 dB) . The correspondingvalues of E , at positions 5 and 8 (off the edges of the plate) are

significantly lower (~15 dB) than at the positions directly under the plate.The E values at positions 1 and 5 are approximately the same as at position

3, indicating the field is relatively uniform along the length of the plate.

The horizontal transverse component, E , for the back plate (figure

4.17) has a total difference in magnitude of approximately 20 dB (± 10 dBfrom positions 1 to 3 and from positions 3 to 6) . This is in good agreementwith the theoretically predicted values shown in table 3.5. Again, E along

the plate, positions 5, 3, and 8 are approximately equal. The E values at

positions 2 and 4 off the edges of the plate are lower than at position 3 byapproximately 2 to 5 dB . The field distributions across the width and alongthe length of the back plate are reasonably uniform in the described testvolume. The cross polarized component, E , however, is almost as strong as

the dominant E .

z

15

The horizontal longitudinal component, E , for the side plate (figure

4.19) has a difference In magnitude between positions 5 and 3 ofapproximately 14 dB (± 7 dB) . The difference In E between positions 3 and

8 could not be measured due to Insufficient probe sensitivity and rf Inputpower. We presume, from the results shown In table 3.5, that thisdifference Is approximately the same as for positions 5 and 3. Thetheoretically predicted difference from table 3.5 Is 18 dB . Variations in

E along the length and across the width of the side plate are relatively

small, essentially In agreement with table 3.5. However, because of theexcessively large differences along the length of the chamber associatedwith rf applied to the side plate, use of the side plate to generate TEMfields is not recommended; and for this reason, a similar plate is not to beincluded in the design for the full-size chamber.

At frequencies above approximately 300 MHz, the differences between theaveraged maximum and average amplitudes approach about 8 dB in figures 4.16,4.18, and 4.20, typical of reverberating operation of the chamber. This is

another indication that the chamber is operating properly in thereverberating mode at these frequencies.

The amplitudes of the field components in figures 4.15, 4.17, and 4.19,at frequencies above 300 MHz, are approximately the same. The totalmagnitude or RSS value of the average field in these figures is

approximately 4.8 dB (a ratio of J 3) greater than the individual components.This will occur if E = E ~ E , and indicates that the field components are

X y z ^

randomly polarized In the chamber. The total magnitudes of the maximumamplitudes in these figures, however, are only about 3 dB greater than theindividual components. This indicates that the components' maximxam valuesare not completely Independent of each other. This is similar to theresults obtained for other reverberating chambers and appears to be inherentin the measurement method. The implication is that if multiple receptorsare used (for example an Isotropic probe whose output is a function of allthree orthogonal dlpoles) in establishing the test field amplitude inside a

reverberating chamber, the results will be biased by either 3 dB for maximumresponse data or 4.8 dB for average response data.

A summary of the spatial field distribution obtained from figures 4.15,4.17, and 4.19 are given in Table 4.1. As mentioned earlier, the E- fieldamplitude in the chamber is Influenced by the chamber's Q. Inserting the 7

probes into the chamber with their lossy transmission lines lowers thechamber's Q and hence the field strength in the chamber compared to whatwould be present without the loading effect of the multi-probe system.

4.3.7 Test Results from 20 GHz to 40 GHz

Measurements were also made to determine the chamber's couplingefficiency (loss) in the frequency range from 20 GHz to 40 GHz. The methodof measurement and the definition of coupling efficiency are the same asthose described in sections 4.3.5 and 4.3.1 except that appropriatewaveguide hardware and Instrumentation were used. The results are shown infigure 4.21, Using these results and K ~ 2.4 in (5), we can predict the E-field to be generated in the chamber by a net input power of 1 W. These

16

results are shown in figure 4.22. Each figure gives the results of two sets

of measurements, one made at NIST in Boulder, CO, and the other at a later

date at USAEPG in Sierra Vista, AZ. The second set of measurements was made

at USAEPG because of the large departure in the first data (between 24 GHz

and 26 GHz) from expected results. The initial reaction to the first set of

data was to question their validity. However, after some research, it was

realized that the data were correct and the differences were due to water-vapor absorption in the frequency range between 20 GHz and 26 GHz as shownin figure 4.23. The curves in figure 4.23 give the calculated normalizedpropagation attenuation/kilometer in air for 5 relative humidities (RH) [17,

18]. The propagation attenuation is also influenced, to a lesser degree, bytemperature and barometric pressure. A check of the weather conditions onthe day the measurements were made at NIST verified that the relativehumidity was high, accounting for the large loss between 24 GHz to 26 GHz.

The relative humidity in Sierra Vista, in contrast, is much lower, about 20

%. These results imply the need to control the environment inside the

chamber, especially for measurement at frequencies above 20 GHz.

4.3.8 Summary of CW Calibration and Evaluation Results

Results of the CW calibration and evaluation of the 1/10 scaled chamberindicate three distinct frequency regions of operation. The TEM mode regionextends from 10 kHz to approximately 70 MHz, slightly lower in frequencythan where the first higher order mode becomes resonant. Test fieldsestablished in this region simulate planar fields with the E- field normal to

the driven plate and the magnetic field tangential around the plate. Largedifferences in the fields exist in the test volume between the plates andground due to the close proximity of the plates to one of the chamber'sinside surfaces. These changes in E-field amplitude vary from ± 6 dBV/mabout the center location for the top plate, ± 10 dBV/m about the centerpostion for the back plate, and up to ± 14 dBV/m about the center positionfor the side plate, for the volume occupying 1/2 the distance between theplates and outside walls or floor. The accuracy with which the E-fieldwithin this unperturbed volume can be established is estimated to beapproximately ± 2 dB. Input power requirements to establish 200 V/m testfield are determined by the plate driven. For vertical polarization (topplate) , about 640 W are required. For horizontal transverse polarization(back plate) , about 4000 W are required. Use of the side plate is notrecommended.

The second frequency region, referred to as the transition region,extends from approximately 70 MHz to 300 MHz. The test fields are a

combination of TEM plus an increasing number of higher order modes, as a

function of increasing frequency. Thus the field is no longer a simpleequivalent plane-wave simulation, but now includes higher order modecomponents which combine with the TEM mode to produce a more complex field.For this reason, the amplitude determination of the test field has errorsthat can be as large as ± 10 dB. (See figures 4.11 and 4.17 for example.)The input power required to generate 200 V/m using the top plate is about200 W at 70 MHz decreasing to about 20 W at 300 MHz. Polarization of thetest field changes from vertical E-field at 70 MHz to undefined(reverberating field) at 300 MHz. The input power required to generate 200V/m using the back plate is about about 1500 W, decreasing to about 20 W at

17

300 MHz. Polarization of this test field changes from horizontal transverseat 70 MHz to undefined (reverberating field) at 300 MHz.

The third (reverberating) region extends from approximately 300 MHz to

40 GHz or possibly greater. The test fields are complex multimoded fields

simulating a near -field environment. The uncertainty of determining the

maximum or average amplitude of the test field varies from an estimated ± 8

dB at 300 MHz down to ± 4 dB at frequencies above 2 GHz. The input powerrequired to generate a maximum test field of 200 V/m varies from about 20 Wat 300 MHz decreasing to about 3 W at 2 GHz, then increasing again to about6 W at 18 GHz. (See figure 4.9 for reference). This discussion is

summarized in table 4.2.

5. Pulsed RF/Time- Domain Evaluation of the 1/10 Scaled TEM/ReverberatingChamber

5 .

1

Background

Parameters of EMI signals that can contribute to upset in electronicequipment include: (a) total energy, (b) peak amplitude, and (c) transientcharacteristics. All these parameters are different inside a reverberatingchamber than in free space. Their characterization inside a reverberatingchamber, particularly for pulsed rf fields, provides information requiredfor making necessary analysis to determine correction factors as a functionof the input pulse parameters. The term "pulsed rf" refers to pulse

-

modulated cw with characteristics similar to a pulsed radar signal. It alsoprovides insight into the inherent limitations associated with using thiscomplex environment for pulsed rf EMS/V testing. Obviously, theseparameters are influenced by the chamber's Q factor, since the time requiredfor a pulsed wave's amplitude to rise to its steady- state value inside thechamber and to decay to zero after the input signal is removed is a functionof the chamber's Q. These charge and decay times can be reduced byartificially lowering the chamber's Q, for example, by inserting smallamounts of rf absorber. However, this reduces the accuracy of determiningthe test field amplitude. Results of work to evaluate the responsecharacteristics of the chamber when excited by pulsed rf of various pulsewidths and frequencies, and with the chamber loaded with a small amount ofabsorber, are presented and discussed in this section.

5.2 NIST Pulsed RF Evaluation System

A block diagram of the system used for the pulsed rf evaluationmeasurements of the chamber is given in figure 5.1. A test field is

established inside the chamber by means of a pulsed rf source connected to abroadband ridged horn antenna. The input signal is monitored by acalibrated diode detector connected to the side arm of a calibrateddirectional coupler. The output of this detector is connected to onechannel of the transient digitizing oscilloscope. The field establishedinside the chamber is monitored by a broadband ridged-horn receiving antennaand calibrated diode detector connected to the second channel of the sameoscilloscope. The receiving antenna is similar to the antenna used fortransmitting the signal. As in the cw tests, the tuner is rotated toredistribute the rf energy inside the chamber.

18

The mode -tuned approach was used when it was necessary to optimize the

measurement accuracy and to obtain complete statistical information for

evaluating the time-domain response characteristics of the signal detectedby the receiving antenna. The mode -stirred approach was used for relativemeasurements with the digitizing oscilloscope placed in its maximum-holdmode of operation. Mode stirring is much faster than the mode tuning

and results in much larger data sampling; however, it provides only maximumvalues. The mode -stirred approach was used to analyze relative informationsuch as charge and decay time and to determine correction factors applied to

the test field amplitude when input pulse widths are too short for the

chamber to charge up to its steady- state or cw- response values. The

digitizing oscilloscope used is capable of measuring signals with rise timesof approximately 30 ps, at a sampling rate of 50 kHz, with sample sizes up

to 1024 per scan.

5 . 3 Pulsed RF Evaluation Results

Measurements were made, using the mode -stirred approach, to determinethe effect of lowering the reverberating chamber's Q on the charge/decaytime of the pulsed rf field excited inside the chamber. Varying amounts ofpyramidal rf absorber were placed in the center of the chamber's test zoneresting upright (cones up) on the floor. Measurements were first made at

900 MHz (the lowest pulsed rf test frequency) , since the response time is

the longest at this frequency. Results of these measurements are given infigure 5.2. The graphs of each figure have two traces. The upper trace is

an envelope of the input pulses applied to the chamber transmitting antenna.It represents the "painted" envelope derived from all the input pulsesduring one complete rotation of the tuner. The signal shown is the reducedvoltage values measured on the side arm of the directional coupler using a

diode detector connected to one channel of the digitizing oscilloscope. Thelower trace is the time -domain envelope of the pulses received by thechamber's reference receiving antenna and measured with a second diodedetector connected to the second channel of the digitizing oscilloscope.The polarity of this pulse has been reversed to separate the two traces.The diode detectors have built-in 50 Q load resistors for impedance matchingto 50 Q. The spread in the top trace (input signal) is due to variations insignal amplitude caused by variations in antenna VSWR (input mismatch) as a

function of tuner position. The magnitude of this variation is frequencydependent, decreasing as the frequency increases as shown in figure 5.3.Figure 5.2 (a) was obtained with the chamber empty (no absorber). Figures5.2 (b)-(h) are for increasing amounts of absorber starting with 1 piece of7.6 cm thick, 61 cm x 61 cm rf absorber, ending with 8 pieces of 30.5 cmthick, 61 cm x 61 cm rf absorber. The effects on both the amplitude andresponse time are obvious. Figures 5.3 and 5.4 show similar results but atselected frequencies up to 18 GHz. Figure 5.3 gives the maximum received rfpulse waveforms for the empty chamber (with no absorber) excited with 5 /zs

wide rf pulses. These measurements were used for determining the chamberrise/decay time as a function of frequency, assuming that the input pulsewidth is wide enough to allow the field in the chamber to charge up to its

steady state value. Figure 5.4 shows the effect of a single piece ofabsorber to reduce the chamber's Q on the maximum received pulse waveform.The input pulse width for these tests was 1 /xs

.

19

It is interesting to compare the chamber's response time calculatedfrom the experimentally determined, average Q values, with the measuredresponse time. The response time is approximately 2Q/w for the signal to

rise or decay to 63 % (1 - 1/e) of the steady- state amplitude. Theseresults are shown in tables 5.1 and 5.2. Table 5.1 gives the measuredminimum chamber losses , the estimated Q factors , and the measuredcharge/decay times determined at 900 MHz for varying amounts of absorber.

The measured rise/decay times were extracted from the results shown in

figure 5.2. Table 5.2 gives the measured average chamber losses, the

associated calculated Q factors, the predicted charge-decay times, and the

measured charge -decay times at the selected frequencies when the chamber is

empty. The calculated and measured response times differ somewhat, but are

still within the margin of error expected for these type of measurements.

Mode tuning was used to accurately determine: (a) the ratio of the

received pulse amplitude versus input pulse amplitude as a function of timeafter the input pulse is turned on, and (b) the time required for thechamber to charge up to its steady- state amplitude and then to decay to zeroafter the pulse is turned off. These measurements were made with the

chamber empty (no absorber) . Results of these measurements are given in

figure 5 . 5 for three input pulse widths , 1 fj-s , 3 fis and 10 /is , and for thesame 10 frequencies. Each graph shows two curves, one for the maximxim andone for the average received signals. By examining these data at theselected frequencies, we can determine the approximate charge-decay times as

a function of frequency. These results are shown in figure 5.6. Similarresults showing the time required for the received signal to rise to 90

percent of the maximum amplitude are given in figure 5.7. The implicationof figures 5.6 and 5.7 is that the input pulse width, at the frequencies ofinterest, should be equal to or longer than the values shown by the curvesfor the pulsed rf test field amplitude to reach 63 or 90 percent of themaximum steady- state amplitude. The steady- state amplitude is achieved ifthe input pulse width is sufficient for the chamber to charge up to 100percent, or to its maximum output for a given input pulse amplitude.

If the transmitted input pulse width is shorter than the charge time ofthe chamber, an error will result in terms of establishing a known peakamplitude of the test signal in the chamber. An estimate of a correctionfactor to apply for this condition can be found by calculating the ratio ofthe received signal amplitude to the steady- state transmitted signalamplitude as a function of time after turning on the transmitted pulse.Figure 5.8 gives these results, determined from the data of figures 5.4 and5.5 with the mode -stirred approach. The data are projected down to inputpulse widths of 0.1 /zs . There is a significant reduction in the correctionby loading the chamber with even a small amount of rf absorber. Figure 5.9shows another way to display the data given in figure 5.8. The correctionfactors are given, for discrete input pulse widths, as a function offrequency rather than for discrete frequencies as a function of input pulsewidth. The symbol, *, indicates the actual data extracted from figure 5.8,and the solid curves are the smoothed approximations for these data.

5.4 Comments on Pulsed RF Measurement Results

Mode -tuned data are generally considered to be more accurate than mode-stirred data for absolute measurements of chamber losses and test field

20

levels . This is so because the measurements are taken for each of the

(stopped) tuner positions and sampled at each instant of time during the

pulse, so that a maximum value (or minimum or average) can be extracted from

the total data taken. Each pulse is digitized at 512 samples in time, for

each of the 200 angular positions of the tuner during a complete revolution.Then corrections are made to each measurement for cable loss, coupling ratioof the directional coupler, attenuator calibrations, and diode detectorreadings according to previous calibration of power response versus signalfrequency and indicated voltage levels. Mode -stirred data cannot becorrected in the same manner; hence, only representative maximum values are

obtained. However, results derived from mode -stirred data usually havebetter definition and are thought to be more accurate for determiningrelative parameters such as the charge time and test field amplitudecorrection factors. This is because larger data sampling approachingcontinuous sampling as a function of tuner position is possible, compared to

the limited (200 tuner position) sampling for the mode-tuned approach.

For part of the data, the maximum amplitude of the pulsed rf fieldstrength, when measured immediately after the pulse turn- on, appears to beslightly greater than the steady- state value. This overshoot is thought to

be due to the imperfect leading edge of the pulse produced by the pin diodeswitch in the pulse source. The overall trend, however, is for the shorterpulses to have a smaller value of peak detected voltage. This is becausethe on- time is not sufficient for the EM energy in the reverberating chamberto reach its steady- state value. Thus, data were taken for a range of pulsewidths to determine the correction factor to be applied when using thechamber to take EMS/V data for very short radar pulses.

6. Projecting the 1/10 Scaled Chamber CW and Pulsed RF Measurement Resultsto a Full -Scale TEM/Reverberating Chamber

The theoretical projected composite quality factor for a full sizechamber (13.1 m x 24 . 1 m x 38 . 7 m) is given in figure 3.8. The K factor,

defined as the ratio of the theoretically determined composite Q by (2) tothe experimentally determined Q' by (3), for the 1/10 scaled chamber, is

shown in figure 4.6 (b) . This factor approaches a constant value ofapproximately 2.4. This value is consistent with K factors determined forother reverberating chambers of different sizes but with similar materials.Thus, assuming the same K factor for the full-scale chamber, we can projectthe minimum loss for the full-size chamber. These results are shown infigure 6.1 for both steel and galvanized steel. The shape of the curve forsteel, at frequencies below 1 GHz, is due to the frequency dependence of thepermeability of steel as was indicated in table 3.6.

The results of figure 6.1 can be used to predict the E-field in thefull-size chamber by using (9) from Section 4.3.3 and the definition of loss

(Loss = P /P_) . These results are shown for steel and zinc coated steel inr t

figure 6.2 (a) when the input power is 200 W. The curves for aluminum andcopper were calculated from the zinc data in accordance with the theory thatthe power density in the chamber is proportional to the square root of theconductivity of the metals used for the wall surface . Thus , aluminum and

21

copper surfaces should result in approximately 1.54 dB and 2.63 dB higherfields, respectively, than the zinc surface of the same chamber size.

Figure 6.2 (b) gives the projected E- field to be generated in the full-

size chamber by an input power of 200 W, based on the data obtained for a

3.69 m X 5 . 18 m X 9 . 79 m galvanized steel chamber located at RADC , GriffissAFB, NY, and for a 3.51 m x 5.18 m x 10.82 m welded-steel chamber located at

the Naval Surface Weapons Center (NSWC) , Dahlgren, VA. A comparison of the

appropriate curves indicates excellent agreement for the steel -chamber data[figures 6.2 (a) and (b) ] but only limited agreement for zinc. Differencesare due to the assumption that the chamber surface is of pure zinc. In

reality, however, the actual chamber coating is not pure zinc.

Figure 6.3 gives the projected charge time for the full-scale, empty,

chamber (no absorber loading) . These results are based on the theoreticaldetermination of Q for steel and zinc chambers and an assumed K- factor of

2.4 for reduction in Q, which is further reduced by approximately 10 dB to

give an equivalent average chamber

.

The projected operational characteristics for the full-scale chamberare summarized in table 6.1. Again three regions of operation are expectedas defined earlier. The results are scaled by a factor of 10 from table 4.2except for the input power requirements for a given E field.

7. Performing EMC/V Measurements Using the TEM/Reverberating Chamber

7 . 1 Performing EMC/V Measurements With Manual Operations

An especially useful application of the TEM/reverberating chamber is toperform shielding effectiveness (SE) measurements. A block diagram of a

simple, manually operated system, which does not require the use of acomputer controller, is shown in figure 7.1. The system uses a sweeposcillator and spectrum analyzer to cover approximately octave bandwidths inthe frequency range of interest. Two measurement approaches are requireddepending upon the region of operation (TEM or reverberating mode)

.

For cable/connector SE measurements , in the TEM mode region ofoperation, a reference cable is placed in the chamber oriented to givemaximum coupling to the magnetic field of the TEM mode. This cabletypically consists of a 72 fi unshielded parallel wire configured, as nearlyidentical as possible, to the cable/connector under test (CUT). Referencemeasurements are made by establishing the test field in the chamber at thefrequencies of interest and recording the power coupled to the output of thereference cable in channel A of the analyzer. The net input power for thesemeasurements is recorded on the Y.. channel of the xy recorder using the time

base output for the x axis. The time base is set the same as the oscillatorsweep rate and is triggered at the same time the oscillator starts itssweep. The CUT is then placed in the same location and orientation as thereference cable and the test field is reestablished at the same testfrequencies. The power coupled to the output of the CUT is recorded inchannel B of the analyzer at the same time the net input power is recordedon channel Y- of the xy recorder. Impedance mismatch errors are corrected

by using (7) as described in section 4.3.1. The SE is then determined by

22

SE = 10 log !£HLlii£ie , (10)

ref i (CUT)

where P^tt't ^^ ^^^ power coupled to the CUT; P. .p,,^,^ is the net input power

applied to the chamber when measuring P^,,„; P ^ is the power coupled to the•^^ ^ CUT ref '^ ^

reference cable: and P. , ,-. is the net input power applied to the chamber1 (ref) r r rr

when measuring P r--" refMeasurements are made in the reverberating region of operation in much

the same manner as for TEM mode of operation. The exceptions are that nowthe tuner(s) are rotated at revolution rates fast (about 1/s) compared withthe sweep oscillator sweep speed (100 s/sweep) . The spectrum analyzer is

set up to measure in its maximum hold mode and tc scan the same frequencyband as the sweeper at fast scan rates (20-30 ms/scan) . Resolution andvideo bandwidths compatible with the scan rate, dynamic range requirementsare selected to ensure that the analyzer is operating in calibration. If

the SE is anticipated to be high (> 40 dB) , a calibrated attenuator is usedbetween the output of the reference antenna and the analyzer. The referenceantenna replaces the reference cable at frequencies in the reverberatingmode region of operation.

The steps for performing the measurements proceed as follows: First,

connect the reference antenna to the analyzer and measure its received powerin dBmW for the fields established in the chamber by the power applied to

the transmitting antenna or plate. Take a sweep and store the measurementresults in one of the channels, for example channel A, of the analyzer whilealso recording the net input power on channel Y- of the xy recorder. Next,

connect the output from the CUT up to the analyzer and repeat themeasurement, being sure the net input power remains the same or is measuredand recorded on channel Y- of the xy recorder. It may require up to 3 or 4

sweeps to fill in the maximum received power across the frequency band.Store the results in dBmW in the analyzer's second channel, for examplechannel B. Then subtract channel B from A and add the calibrated attenuatorfactor in dB, if used, to get the SE in dB . This assumes that the cablelengths between the reference antenna and the CUT to the analyzer are thesame. If limited sweeps are made, use the envelope of the CUT data to

subtract from the reference data so that the minimum SE values are obtained.Using this procedure will ensure adaquate accuracy without requiring longsweep times or a large number of repeated sweeps. (Normally it is notnecessary to wait for a maximum recording at all frequencies.) These datagive nearly continuous frequency coverage measurements of SE. Measurementsof EUT response to determine the susceptibility or shielding effectivenessof enclosures proceed in much the same manner as above.

For some applications, it may be advisable to determine the shieldingfactor (SF) of the EUT. This is defined as the ratio of the EUT's responseexpressed in power to the test field's power density,

SF = 10 log (Peu^/P^') , dB/m2 (11)

23

where P^Tirr ^^ ^^^ power coupled to the EUT in W, and P ' is the powerEUT O-

density inside the chamber in W/m^ . The SF is similar in concept to the

well known antenna factor used in antenna calibrations.

An alternate approach to these measurements is shown in figure 7.2.

This block diagram gives the setup for measuring radiated emissions or SE bydriving the CUT in the radiating mode or by measuring the radiated emissionsfrom the EUT. (For radiated emission measurements, the EUT is not connectedto the rf source, but, rather, becomes the source itself.) For SE

measurements , mismatch errors caused by large VSWRs of the CUT are minimizedby using the bidirectional coupler to measure the net input power to the EUT

and to the chamber's transmitting antenna, while measuring the powerreceived by the reference antenna for each case. Use of a bidirectionalcoupler to measure the net input power gives the equivalent of an impedancematch source for the measurements. For the case of cable, the SE is thendetermined by

SE = 10 log'n(CUT) ^ref(t)

^^ ^^2)

^n(t) ^ref(CUT)

where P ,^^_, is the net input power to the the cable; P j~/^TTrr.x is then(CUT) ^ ^ ref(CUT)

power received by the reference antenna when power is applied to the CUT;

P , . is the net input power to the chamber's transmitting antenna or plate;

and P r-,^. is the power received by the reference antenna when power isref(t) r y r

applied to the chamber's transmitting antenna.

For the case of an EUT in general, the power radiated from it excitesthe chamber and is coupled to the reference antenna. This coupled power is

measured on one channel of the analyzer. The EUT is then turned off andpower is applied to the chamber's transmitting antenna or one of the plates.The output power from the oscillator is adjusted to give approximately thesame coupled power to the reference antenna. This may require the use of a

calibrated attenuator between the reference antenna and analyzer to keepfrom reducing the oscillator output power too low. These measurements arerecorded on another channel of the analyzer. The radiated power from theEUT is then determined by:

= P^^^f(EUT)

3rad(EUT) n(t)

'^ ^

ref(t)

where P j~,„TTmN is the power received by the reference antenna when the EUTref(EUT) ^ '

radiates

.

An example of SE data obtained using the block diagram of figure 7.1 is

given in figure 7.3. The EUT is a TEM cell having a cross -section area of 3

cm x 6 cm with a 15 mm circular aperture. This EUT is a control standarddevice developed at NIST for SE measurement standardization. Two traces areshown in figure 7.3. The top trace is the measurement of the reference

24

antenna received power recorded, in dBmW, on channel A of the analyzer. The

bottom trace is the received power coupled to the EUT recorded, in dBmW, on

channel B of the analyzer. A 10 dB attenuator was used between the

reference antenna and analyzer in performing this measurement. This 10 dB

must be added to the difference, P ^ - P„-_, , in calculating the SE of theref EUT °

EUT.

7 . 2 Automated EMC/V Measurements

7.2.1 TEM Mode Measurement Procedures

Detailed step -by- step measurement procedures using TEM cells for EMCmeasurements are given in [1]. Measurements using the combined chamber in

the TEM mode region of operation follow the same steps. These consist of:

(a) Determining the EUT test parameter requirements in terms of testfield amplitude, waveshape, modulation, frequency range and interval,exposure polarization, exposure time requirements, and EUT performancedegradation criteria;

(b) Placing the EUT inside the cell in one of two locations fortesting. The first location is close to the ground plane to minimizeexposure of the EUT's input/output, power, and monitor leads to the fieldgenerated inside the cell. The second position is midway between the plateand ground plane. This position provides greater field exposure to the

EUT's leads and hence gives an indication of how energy is coupled to theEUT;

(c) Accessing the EUT as required for operation and performancemonitoring. The EUT input/output and ac power leads should be as nearly thesame as in its intended use. Leads should be the same length, if possible,and be terminated into their equivalent operational impedances so as to

simulate the EUT in its operational configuration. Care must be taken inrouting the leads including monitor leads (if not transparent to the rffield) inside the chamber for obtaining the most meaningful, repeatableresults. Placement of the leads, wiring harnesses, etc., should be recordedfor future reference so that if repeat measurements are required, the leadscan be placed as nearly identical to the original tests as possible. Fiberoptics or high- resistance lines should be used for monitoring the operationof the EUT if possible. Such leads minimize coupling to and distortion ofthe test fields;

(d) Connecting the measurement system as shown in figure 7.4; and

(e) Applying power to the chamber's TEM lines with the EUT in operationand recording the EUT's monitor responses as a function of the test fieldlevel, modulation parameters, polarization, and frequency to determineimmunity.

7.2.2 Reverberating Mode -Tuned Measurement Procedure

A detailed step-by-step measurement procedure and flow diagram for

performing automated radiated susceptibility tests using a reverberating

25

chamber is contained in [5] . The procedure is the same for the combined

chamber used in the reverberating region of operation. It is summarized

briefly as follows:

(a) Determine the testing requirements for the EUT as referred to in

section 7.2.1 item (a) above;

(b) Place the EUT inside the chamber and access it for operation and

monitoring;

(c) Connect the measurement system as shown in figure 7.5;

(d) Input the measurement and calibration parameters into the computer.

These include frequency ranges and increments for the tests, number of tuner

steps to be used, signal generator output levels to generate the requiredtest field amplitudes, EUT output level maximum limit to prevent damage, EUT

response time, and cables, directional couplers, and sensors calibrationfactors ; and

(e) Perform the measurements in automated sequence starting at the

lowest frequency and level, increasing the frequency in increments at a

constant exposure level and then increasing the levels until the maximumtest level is achieved or the EUT maximum allowable response is reached.

Correct the measurement data using the appropriate calibration factors

previously put into the computer. (This can occur simultaneously with makingand recording the measurements.) The optimum technique for obtaining mode-tuned data is to step or increment the frequencies at each tuner positionbefore stepping the tuner to its next position. When the tuner is stepped,

it should be allowed to stabilize while power meters, voltmeter, monitors,etc., are being zeroed and residual offsets measured. Obviously the rf is

turned off while this is done. The rf power is then turned on; the testfield is established; the measurements are made at each frequency; and thedata are corrected (with respect to a given input power) and stored. Thecorrected data are then configured at the end of the run, after the tunerhas been stepped through all positions to complete one revolution (360°) andcompiled by frequency. Various calculations of averages are then made andthe data are presented in the final form.

7.2.3 Reverberating Mode -Stirred Measurement Procedure

Although the mode -stirred measurement approach begins in a similarmanner to the mode -tuned described above, the automated measurement sequenceof events is somewhat different, and hence some parameters necessary forperforming the measurements differ from the mode -tuned approach. Theprocedure is summarized briefly as follows, after determining the EUT testparameters

:

(a) Place the EUT inside the chamber and access it for operation andmonitoring;

(b) Connect the measurement system as shown in figure 7.5;

(c) Input the measurement and calibration parameters into the computer.

These include the frequency ranges and increments, the signal generator

26

output levels , range and increments , the EUT output level minimum and

maximum response limits, the EUT response time, the cables, directionalcouplers, and sensors calibration factors, and the tuners rotation rates;

(d) Perform the measurements using the automated sequence, whichincludes setting the generator to its first frequency, starting the tunersin motion, and monitoring the EUT response (s) while slowly increasing the

test fields. When the EUT minimum response is exceeded or the maximum testfield is reached, the tuners are stopped; the system is initialized with the

monitor voltmeters; and the spectrum analyzer is set up to start taking datawhen triggered. The rf power is turned on, the net input power is measuredand the tuners are placed into motion as the voltmeters and spectrumanalyzer are triggered into performing their measurements;

(e) Make calculations of averages and corrections for calibrationfactor while the measurement proceeds. The results are then Q,<5mpiled at theend of each frequency run after one complete revolution of the tuner andstored in memory for presentation at the conclusion of the measurements. Ifthe maximum EUT response is reached during the measurement, the generatoroutput is reduced by an interval of 1 dB and the measurement is repeatedstarting with a complete new initalization and a full 360° revolution of thetuners. This ensures accurate results without damaging the EUT.

7.2.4 Comparison of TEM/Reverberating Chamber Measurement Results withAnechoic Chamber Tests - Some Examples

Comparisons of the EUT response obtained using a reverberating chamberwith anechoic chamber data are typically made in terms of maximum or peakvalues. The main reason for this is that the EUT's worst- case performanceor susceptibility is normally what is desired. An additional practicalconsideration is that it is very difficult to obtain a true average EUTresponse from anechoic chamber data. Even determining the EUT's peakresponse in an anechoic chamber may require considerable effort involvingcomplete pattern measurements

.

The benefit of making this type of comparison is to obtain a"correlation factor" between results obtained in the TEM/reverberatingchamber and those obtained in an anechoic chamber. This was done for threereference standard EUTs . These were: (a) a 1 cm dipole probe, (b) a smallbroadband horn antenna similar to what is used at frequencies from 0.8 to 18GHz as the transmitting and receiving antennas in the chamber, and (c) asmall TEM cell with 3 cm x 6 cm cross section and a 15 mm diameter circularaperture. The measurements were performed in a combination of facilitiesincluding the RADC's small 1 . 41 m x 1 . 57 m x 1 . 75 m aluminiam reverberatingchamber located at Griffiss AFB, NY; the NIST 2 . 74 m x 3 . 05 m x 4. 57 m steelreverberating chamber located in Boulder, CO; the USAEPG 1/10 scaled 1.31 mx 2.41 m X 3.87 m TEM/reverberating chamber located presently at Ft.

Huachuca, AZ; and the NIST 4.9mx6.7mx8.5m anechoic chamber located inBoulder, CO.

The maximum responses obtained using the 1 cm dipole probe measured inthe 1/10 scaled TEM/reverberating chamber, the NIST reverberating chamber,and the NIST anechoic chamber are given in figure 7.6. The probe responsein the anechoic chamber is greater than its response in the reverberating

27

chamber and TEM/reverberating chamber by an average of about 2 dB. This is

expected, since 2 dB corresponds approximately to the free-space gain of an

"electrically small dipole. Thus, the correlation factor between the

reverberating chambers' and the anechoic chamber's results is the free-spacegain of the EUT [5]

.

Measured maximum responses of a broadband horn antenna obtained usingthe 1/10 scaled TEM/reverberating chamber and the NIST anechoic chamber are

shown in figure 7.7. The horn was aligned on the boresight (main beam) ofthe transmitting antenna for the anechoic chamber measurements to obtain the

maximum response. As expected, the horn response is greater in the anechoicchamber than in the TEM/reverberating chamber by approximately its free-

space gain. This is shown in figure 7.8 where the horn's calibrated free-

space gain has been subtracted from its measured maximum response. Clearly,the results are in general agreement.

^ Figure 7.9 shows a comparison of the shielding effectiveness of a small^apertured TEM cell obtained from theoretical calculations, from measurementsmade using the RADC small reverberating chamber, and from measurements madeusing the 1/10 scaled TEM/reverberating chamber. Both mode -tuned and mode-

lstirred approaches were considered using the scaled TEM/reverberatingbhamber. The RADC chamber measurements were performed using the mode- tuned

^approach. The mode- tuned results for the two chambers are in goodagreement, but are greater by a few decibels at the higher frequencies thanthe theoretically determined values (straight line). The mode-stirred dataagree better with the theoretically determined values. The limited numberof tuner positions (200) used for the mode-tuned measurement may explain thehigher SE results, since a limited sample at higher frequencies does notallow a true measurement of the peak response of the EUT. In general,however, the results agree within predicted measurement uncertainties.

8. Svumnary and Conclusions

The designed hybrid chamber has three distinct regions of operation:(a) TEM mode, (b) transition, and (c) mode-stirred. For the 1/10 scalechamber, the TEM region covers the frequency range from approximately 10 kHzto 70 MHz; the transition region covers the frequencies from 70 MHz to about300 MHz; and the mode -stirred region covers the frequency range from 300 MHzto > 40 GHz. In the transition region, the first few higher order modes areexcited inside the chamber, but in insufficient number for the chamber to

reverberate properly. Thus, the test field is a combination of the TEM modeand a limited number of higher order modes that yield a complex field thatis difficult to accurately determine both in polarization and amplitude.Techniques such as absorber loading, to suppress the Q of the higher ordermodes, and test field averaging over the test volume, using multiple probes,may be helpful in reducing errors in establishing the test field and indefining the test field parameters.

Test fields are excited in the chamber at frequencies up to

approximately 1 GHz by using the transmission lines. At frequencies aboveapproximately 1 GHz, ridged horn antennas are used. The parallel platelines should be unterminated (open circuit) with the bulkhead feed- throughconnectors capped to prevent rf leakage to outside the chamber when they are

28

not in use. Other important characteristics may be summarized as follows,

together with some recommendations:

(a) Spatial variations of the maximum and average electric fields in

the test volume of the chamber are given in table 4.1 at a few selecteddiscrete frequencies. These data were obtained using the NIST multiprobesystem and the mode -tuned approach with 200 tuner increments per revolutionat frequencies from 0.03 to 2 GHz. The limitation for determining the

spatial E-field variation depends on the increasing mode density and hencefield complexity in the chamber as a function of increasing frequency. Thespatial E-field variations should decrease as frequency increases above 2

GHz to less than the result of ± 2 dB shown in the table for 2 GHz.

(b) Expected field levels to be generated in the three frequencyregions by a given input power are presented in table 4.2, together withestimated uncertainties.

(c) The maximum E-field established inside the chamber at frequenciesabove 300 MHz is about 7 to 8 dB greater than the average E-field. This is

consistent with the previous results obtained for other reverberatingchambers such as the RADC large chamber , the Naval Surface Weapons Center(NSWC) chamber, and the NIST reverberating chamber.

(d) Reference standard EUT responses determined in this scaled chamberare approximately the same for the same EUT at the same field levels as

using RADC, NSWC and NIST reverberating chambers.

(e) The directional characteristics of an antenna or EUT placed insidethis hybrid chamber operated in the reverberating region of frequencies arelost, resulting in an equivalent gain of dB in the highly complex fieldenvironment.

(f) The correlation factor between free-space (anechoic chamber) andreverberating chamber measurement results for the same EUT is the free -space(far-field) gain of the EUT. This implies that susceptibility criteriadetermined for an EUT using a reverberating chamber should include anadditional factor proportional to the EUT's open-field estimated gain as a

function of frequency.

(g) If the chamber is used for pulsed rf EMS/V testing, and the inputpulse widths are shorter than the chamber's charge time, an error willresult in establishing a known maximum amplitude of the test signal. Anestimate of the correction factors to apply for shorter pulses is given infigures 5.8 and 5.9.

(h) The use of rf absorbing material inside the chamber significantlyreduces the Q and charge time of the chamber, and hence should be consideredas an effective means to improve the fidelity of the test pulse waveform to

be applied to the chamber. However, care must be exercised to ensure thatthe absorber would not excessively reduce the Q so that the referenceantenna received power drops by more than 6 dB , or the tuner effectivenessis reduced to less than 20 dB.

29

(i) The use of rf absorbers also reduces the amplitude and variations

In the E- field generated In the transitional frequency region, and thus the

uncertainly In establishing the test field level, at perhaps the expense of

Increased rf Input power requirements

.

( j ) The mode -tuned approach Is recommended for use at frequencies from70 MHz to 2 GHz with at least 200 tuner positions per revolution, if

absolute levels of the test field are required. The mode -stirred approach

is recommended for use at frequencies above 2 GHz for absolute measurements(with more than 3000 samples per revolution for frequencies between 2 to 4

GHz, and more than 5000 samples per revolution for frequencies above 4 GHz).

The mode-stirred approach is also recommended for all frequencies above 300

MHz if only relative measurements are made.

(k) Antennas for transmitting rf energy into the chamber and for

measuring the test E- field should not be used outside their specifiedfrequency range. For example the 0.8 to 18 GHz ridged-horn antennas shouldnot be used outside their specified band.

(i) Additional work is needed to verify experimentally the correctionfactors shown in figures 5 . 8 and 5.9. This could be accomplished bycomparing the measured responses of a well -characterized EUT to cw andpulsed rf fields as a function of frequency and pulse widths using both ananechoic chamber and the TEM/reverberatlng chamber.

(m) If corrections are not made for mismatches of the transmitting andreceiving antenna, absolute amplitude measurements of the test field insidethe chamber will be lower than the actual values. This will cause an errorin determining the EUT response, which will also be too low. Experiences inperforming this type of measurements indicate that these errors may be as

great as 5 dB when VSWRs exceed 15 to 1,

(n) The average wave impedance in the reverberating region ofoperation, when the maximum response of an EUT is measured, appears to behigher than 377 Q. This is especially true for frequencies lower than 2

GHz. This means that if a wave impedance of 377 Q is assumed whendetermining the maximum amplitude of an exposure field, there will be a

systematic error resulting in too low a calculated E-fleld, and thus also intoo high an EUT response. If the field strength is determined by using a

calibrated E-field probe, there is still some uncertainty, since theenvironment in which the probe is calibrated is different than thereverberating chamber environment.

9

.

Acknowledgments

Work described in this report was sponsored by the U. S. ArmyElectronic Proving Ground, Ft. Huachuca, AZ , with Robert Weeks as projectofficer. The authors acknowledge his assistance in making this projectpossible. In addition, the authors express their appreciation to MotohisaKanda for his support and to J . W. Adams for his helpful comments.

10. References

[1] Crawford, M. L. Generation of standard EM fields using TEM

transmission cells. IEEE Trans. on Electromag. Compat. EMC-16

(no.4) : 189-195; 1974, November.

30

[2] Crawford, M. L. ; Workman, J. L. Using a TEM cell for EMC measurementsof electronic equipment. NBS Tech. Note 1013; Revised 1981 July.

[3] Corona, P.; Latmiral, G. ; Paolini, E. ; Piccioli, L. Performance of a

reverberation enclosure for power measurements in the microwave range.

2nd Symp . Tech Exhibition on EMC; 1977; Montreux, Switzerland.

[4] Bean, J.L.; Hall, R.A. Electromagnetic susceptibility measurementsusing a mode -stirred chamber. IEEE Int. Symp . on EMC; 1978 June;

Atlanta, GA.

[5] Crawford, M.L.; Koepke , G.H. Design, evaluation and use of a

reverberation chamber for performing electromagneticsusceptibility/vulnerability measurements. NBS Tech. Note 1092; 1986

April.

[6] Ma, M.T. ; Kanda, M. ; Crawford, M.L.; Larsen, E.B. A review ofelectromagnetic compatibility/interference measurement methodologies.Proc. of IEEE; Vol. 73, No. 3; 1985 March.

[7] Liu, B. H. ; Chang, D. C; Ma, M. T. Eigenmodes and the compositequality factor of a reverberating chamber. NBS Tech note 1066; 1983Augus t

.

[8] Dunn, J. M. Local, high-frequency analysis of the fields in a mode-stirred chamber. IEEE Trans, on Electromag. Compat. EMC-32 (no. 1):

53-58; 1990 February.

[9] Corona, P.; Latmiral, G.; Paolini, E. Performance and analysis ofreverberating enclosures with variable geometry. IEEE Trans, on EMC;

EMC -22 (no. 1): 2-5; 1980 February.

[10] Wilson, P. F. On slot-coupled waveguide and transmission linestructures. Ph.D. Dissertation, Dept. of Electrical Engineering, Univ.of Colorado, Boulder; 1983.

[11] Crawford, M. L. ; Koepke, G. H.; Ladbury, J. M. EMR test facilities -

Evaluation of reverberating chamber located at RADC , Griffiss AFB,Rome, NY. NBSIR 87-3080; 1987 December.

[12] Crawford, M. L. Generation of standard EM fields for calibration ofpower density meters: 20 kHz to 1000 MHz. NBSIR 75-804; 1975 January.

[13] Bowman, R. R. Calibration techniques for electromagnetic hazardmeters: 500 MHz to 20 GHz. NBSIR 75-805; 1976 April.

[14] Larsen, E. B.; Ries, F.X. Design and calibration of the NBS isotropicelectric-field monitor [EFM-5], 0.2 to 1000 MHz. NBS. Tech. Note 1033;

1981 March.

[15] Kanda, M. ; Driver, L.D. An isotropic electric-field probe with tapered

resistive dipoles for broad-band use, 100 kHz to 18 GHz. IEEE Trans,

on Microwave Theory and Tech. MTT-35 (no. 2); 1987 February.

31

15] Bensema, W.D.; Reeve, G.R.; Koepke , G.H. A multisensor automated EMfield measurement system. Proc . IEEE IMTC ; 1985; Tampa, FL.

17] Liebe, H. J. An updated model for millimeter wave propagation in moistair. Radio Science 20 (no. 5): 1069-1089; 1985 May.

18] Liebe, H. J.; Layton, D. H. Millimeter-wave properties of theatmosphere: Laboratory studies and propagation modeling. NTIA Report87-224; 1987 October.

32

E•f—

-I-'

03S-O)X!i-

>S-

-o

oC/)

oLlI

<en3

w0)

au0)

M

1—1

C8

co•^h^

a<u

«i >

^4

0)

w JDw c

CO

u J=u (J

CM

to

fa

33

34

£•r-

-M«as-o;

i-

>(Us-

T30)

<oOCO

0}

uo

c 1^O 00

•r-l

•u m

J-i E

W t-l

i-( <!•

•O •

CM

o ee

i-i

mr-l •

o ^^iH•U J-l

(U OJ

S-i ^o e

Eh O

CO

m

CD 00 |\ CD lO -^ rO

(%) dB9 epow

CnI

35

2b

2w

Center conductor^

s

Outer conductor

/

2a

Figure 3.4 Cross-sectional view of a TEM cell for

determining the field distribution.

E^ = z sin(mTT/2) cos(rmTx/2a) [Jo(g') + ^7(9') + •.•]-y mo

^

E = : sin(mTT/2) sin(m7Tx/2a) [Jo(g') + Jo^S') "^ •••^

cosh 2^(h. - y)

sinh 27 h^-

Si'nh g (hi - y)

sinh -^ n.

where mQ = 1 , 3, 5, . .

.

andg ' = m7Tg/2a, g = a - w.

36

Perssnneldoor

24. 1^

Rnteroom

BULKHERD-^^

75 Ohm ^I

Plate Line ^

,b.

TUNER

BULKHETIDTO 75 Ohm Termination

k 8.88

1

:i>

TEST VOLUME.LWE/_

/X.

38.7 m

BULKHEHDTO 75 Ohm Termination

B.88"I

,b.

I TUNER

e.ee

13. 1 m

^ '"^^

TEST VOLUMEBm X 16m X 3Bm

\l\

HRNGRR DOOR - (SHIELDED)

/

Figure 3.5 Cross - sectional sketches of a proposed full scale TEM/reverberatingchamber (13.1 m x 24.1 m x 38.7 m)

.

37

oo

oO)

0)

T-l

cd

om

oCO

ft

0)

wo •

( ) CL^r--

^_^

O EU

00N CO

X 0)

i-i

6o >03 ^-^

u •

O <f

>N 4-1 CM

u X

c C E

Oin

o•H .-1

4J •

3 CO

U •r-l >WS-l

•P >-l

s^ CO 0)

1 1 , •O E

o

al

mode

berating

(

oTheoretic TEM/rever

o ^Osl CO

0)

3)

ofo

LO -vh ro

(%) dea apow

CnI O

38

N

1u

c•H4-)

0)

>

M\sUH0)

o '^CO ^ r-l

CO

U J-l

O

C *4-l

om(D c

o

U P3

(D•r-l

L_ )-l

h00en

(U

•o eo

l-l COcfl

o X•r-l

(V

uo0) vO

CM

m

s-i

%) deg apoiAi

39

TTTTT iiiii Mil rnrrm—r

liii till

iini i I I I mil Mil tmt I I I I MiM I I I I

QCVJ

bO y^S

c o•H c4-) • r-l

« NVJ v,-^

Q>

^ "O!-l (D

OJ N> •r-l

(U cu cc\ >s i-i

u (t

H tc

(B ,^i-H tfl

C5 \_'

O I-l

W .. (1)

i-t C 4->

OJI—

1

Eiw 3 ^T5 w <u

Q) c3 .—

1

W I—

1

o ^ oa E !-i

<-—

^

oN u r-- T3=C a • I-lo 00 o

' '

<+-i ro Oo x.^

>1 Xu -0E r-i S 0)O) o u3 4J r-l Cc o • -rH

OJ C3 -^ CCS- iw CM a.u.

I—

1

>^ X <U

4J E1-1 E

DJ 1—

t

2E TC

C• 13 0)

Cr r-lQI-l i-^ ^^cC >>•

o TJT-l cU ^^ CS

a; 0)

Vj ^ --1

o E 0)

a; cC <D

.c ^ -i-i

H o w

00

en

0)

j-i

3bO

•r-l

tn

DJ

QG) Q

Q^OL X >>»o:^De^ ^ ar^LsodujOQ

40

oO) .

c -l-> ,—

V

1

—

03 4-> CQOJ •r— o<u c E+J •1

—

C/1 CO</1 E

+J n3 +1^—

*

rt3 t. -s

-C .E 4->

to -M CO•r— OJ •r—

E -E E•1

—

+-> rn s-M- •1— CO O)

2 -QCJ s- Ec Vr o ro•I— M- -EN CO

<UU

CTi U uQ E •I—

OJ X (Ti oIM (-> x:

"r

—

E CO uc •1— cufO CM -o E> . fOr— LD EfO O ECT X

+j"

fC F fO Es- o

c r~>- (T3 • 1

—

•1— . Q-+->CO OJ •I—

CO CO E-o M- • 1—1

—

O E M-O) CD.r- i- co "O<4- QJ

-Q 4-> (D+-> E fO ECO fO 3Ol -E S- r—+-> U <u o

J3 >M- CD EO E fO -M

•r— jz: COE -M u OJO ro 1—CO i. u•1— O) •r—

S- J3 o .

(T3 S- -E s-Q. O) O OJE > OJ 2O 0) E o<_> s- fO Q-

CO

E

toEo(/)

EO)_E

Q

CO "^CO CT)

CO C3^ • •O ^ p— OCO r- + 1

-H +1 CO CO

II II II II

I— CMX ^ rvi ivj

Lf) enCM 00

I— ^ .o ^ .— oCO .— + 1

+ 1 +1 CO CO

11 II II 11

1— CMX >- rvj r-vi

"vT 00CM 00O CO • .O "^ .— o

CO I— + 1

+1 +1 CO CO

II II 11 11

1— CMX >- l~vl M

«^ COCVJ 00O CM • •O <* r— O

CO I— + 1

+ 1 +1 CO CO

II II II II

1— CMX ::^ M M

CQ-o

CL.

--DQ_

cCM

CO

CM

cn

CM

LO

CDCO

Frequency,

MHz

oo ooCM

1

ooCO

ooID

41

E OJ E•I- -M-M +-> CO

s- E -ocu to cS- roOJ S- .—> +->

O) II

S- O)E S-

r— 03 O)<L) LO JDOJ E4->

C/l

ro

Ou

N O)•1- JDc EfO 03> -EI— UroCT> Uro oc O (/)

•I- O) <uE O

CO ro E"O ro1— E -!->

OJ ro </)

M- E X5

+J EOO -t-J OO) ro •-+j jr +J

+-> roH- s-

o x: ro+j a.

E -c- O)o S OO</)

•I- i-

ro -QCl eE rooo c_>

cu

o

CO

ro

roEEO)+->

E<

CQa

-o

--0Cl.

CQ-a

s-O) »>

-MO) -Ds: D-

I— -X!

00

C\J

OO

CO

CMCM

CO

oCM

CO 00 00

LT)

00 00

OO

OO

OO

C\J CM

00

CM CM CM

IT)

CM

-aCTi CM un ID ir>

CM

n(U

ro

oE0)

,3 IM CM ID o c o o o o

, cr n: • « • • • • " •

' OJ CD o o 1

—

CM 5d- 00 CM COs- r-^ 1

—

' u.

.

42

E I/)

O 1—

•r—+-> EfO OS- •r-

(T3 +JCL •1—

QJ EVI •r—

E O)

roQJD E

c 3'^

"oE >--

+-> O)(0 1—

S- EO) • S-JQ E CL)

F •1— +->

fO (T3 +->

-C cn fOu

toa.

o E EE O

o O) •r—

.c -M -»->

u E (OCD (O •r—

E T3fC cn (O

E i-

E •r—+J OJ+-> x:

(/) •1— 4J-o E

to M-cu E O•I— mM- s- to

+-> +J-a E<D -a •1

—

+-> 0) OfO -i-> Q.•r— u-o O) S-fO 1

—

O)s- O) s

to os- Q.o -E 1

q; 5 ro

3 to1

—

<U Eo o O> E

ra "O-M -!-> OJto to to

O) •t— fO1— -a JD

ro

ro

0)1—.

J3<o

toE "^ CO ^ CO «;:f 00o

r-- CM 00 OJ CM 00 ivl- CM CM 00•*

00 r^ • • r^ 00 • • CO CO • •

in• I— . . ,— • • 1—

E E I— CO + 1 00+ 1 00+ 1

E +1 +1 CO ro -H -H CO CO +1 41 CO COII II II II II II II M II II II II

•*

CM 1— CM I— CMQ X >- rCTrvJ X >- M M X >- M M

M-E OCO E <-< CM LO

cn CMII -n- CVJ

to Q CMa: E

1E

^^ -O E r^ CT>•1—

• " "^"^^ un 00 I

—

c x: 1— cvj • • •

''~fO :z) Q 1

—

% s: LU CVJ

Qto CM CO r— Cr> 1— <T> r— CTl

<U E <>D CM * C\l "^ r-- ^ CM r— I— «^ CME O C>J CM • • r— 1— • •

3 •1— 1— • • . .

r—

'

to +i +\ + 1 00+ 1 00+ 1

o E E +1 HH 1— 1— + 1 +I1— -—>

1 ^ II II II II II II II II II II II II

E CM 1— CM I— CM+->

to

1—

Q X >- PvJ M X >- M M X >- rM M

1M-OE .-C CO Lf) <£>

E O --^ CM tn >—

•r- CMI— to Q , CD

EII O)

Ed; -r-

-o E ^• r> "''*»»^ r~~ cr> r—X 1— CM CO CO 1

—

(O :z) Q • • •

2: Lu osj

1

c

x:(->

T3•I- CU OJ

S E O)

E ro S-fC 1— CD ^ p>^ COOJ O- O) to CM f""

J3 1 T3ms-

O)

5O O) OJCL E O)

i fC S- C>J CM14- I— cn CO '

—

.— O- OJ I

—

fO 1 "Oin Lu

na E E•I- O) ro cnE +J -a • f

•

E 1^ LO '~~

2: <: ' CM

>>uE<U N CM13 3: • • •

crtD r— CMCU

""

s-Ll_

43

Table 3.4 Theoretical mode calculations of frequency and gap betweenadjoining modes for 1/10 scaled TEM/ reverberating chamber,

a = 1.31 m, b = 2.41 m, and d = 3.87 m.

Mode Indices Frequency Gap Size# MHz MHz

1 Oil 73.36 73.362 012 99.43 26.073 101 120.80 21.374 110 130.27 9.475 021 130.47 0.206 013 131.89 1.427 111 135.92 4.038 102 138.19 2.289 022 146.72 8.5310 112 151.58 4.8611 103 163.11 11.5312 014 167.05 3.9413 120 169.15 2.1014 023 170.40 1.2415 121 173.53 3.1416 113 174.60 1.0717 122 186.06 11.4618 031 190.85 4.7919 104 192.65 1.8020 024 198.86 6.2121 032 202.31 3.4522 114 202.47 0.1723 015 203.52 1.0424 123 205.25 1.7325 130 219.12 13.8726 033 220.08 0.9627 131 222.52 2.4428 105 225.01 2.4929 124 229.43 4.4230 025 230.35 0.9231 201 232.09 1.7432 132 232.42 0.3333 115 233.47 1.0534 210 237.16 3.6935 211 240.30 3.1436 016 240.70 0.3937 202 241.60 0.9038 034 242.79 1.19

39 133 248.05 5.26

40 212 249.50 1.45

44

o oo oo oo oo o oo o oo o oo o o

o o oo o oo o oo o o0(0o oO O IIo o

d d >>

CO ro^ CD

CO vo CDcj c\j o>o in r^to I— o

CO

r^ I—

f— If)00 O)

II

O >>

C\J

<M CVJ

CM 00 00O^ to CM _O r- f— "5

rs. CM I— II

>,

o^ r>» ro ro CMo «o r^ ^ 00Ot r>» (O «£> roo» o 00 CM M

O ^ O CM VO ^^r^ r^ O m cr><r> r>. o O to II

r^ ^ O oo I—>>

1^ ro I— o O

^ CM ^-c •<- •.- •.- c^ £ £U CO

N

in 00 r^ O C3 nr" CJ> «o VO O "00 o CT> CM ° 11

<J» to CT» o O II

o CM a\ ir> om ID ^ CVJ o»» rv VO CM or^ <D CO o o

«o \or- Oo \o

00

CO 00O f-r^ o

CM

CO <IJ-

VO oVO 1^

VO O0> CO

r- I— O

o oo oo oo o

oooo

oooo

CVJ

II

>1

CJ

>>

o ooO IIoo

ca>coa.eo

0)

Eo

o ^^ oO fO Q.O II EO >, o

o<^ ^ •-

O fOO II E

en^ CO o VO 00 ^^CO 00 o o\ CO Its

CO f>» o m r* CM«• CM in o o II

CO V o o d >>

V

CM

MC

^ I-* r^ CO VO c:

<7t o r>. ^ a\ rc o00 CO O 00 Q.

o r>. o ol[

t• • O

CO VO o o d >,

n

uX

in <»» o CM o "^,^ Oli*« ^— VO CO ^ fO«» VO VO o CM ^—r<. CO 00 O

II <a

CM1^

r>. O o o >^ o

oooo oCO or-» o•a- o

CM 1^ o

<uc

4-1

ce

u(0

pa

00 •—

•^ "r- •»- O

0000CMm

oso

m ooooCO00

00 -- O o

o o o ^ o n<r> CM VO ^5 o ^^CO VO VO t^ o ro

fO CO 00 r— o CMII

^_ !>. o O o1^

CM

+->

C0)

VO CO ^— o o ^^ oin r— r^ CM o oo^ a\ o 00 o E

oo o> •>> o o II

o in o o o >, oin

r— inf— CMVO CMa\ CM

CT>

ooomo00ino

8 '^oO II

T301

Ooooo oo oo oo o

oooo

>,

o ^^

8 M

o• >>o

CDI— 4Jf— (Oa; >—U Q-

s: 0)Lu -o1— •>-

(/)

-oQJ—

^

r- J3(O^—

-

OI/) X)c

,—

,

O fO

<u^—

c -~- Ol

(C f— +JfO

o. <u .—x: Q.

X +j>, .:^

c o•1- ro

4-)

Q C (0

(U

(A

o «Q. Ol

E Oo c(J to

-a-a ojI— CL^ <U E•r- •!—11-

1 u(_) -1-

•>- +->

S- CO+-> •!-

u s-<U Q)r— +->

(U O(O

-o s-O) <oN ^•I- (J^—fO c^E1- oo o'- '

LD

ro

OJn~XI<a

•^» -r— 'r* OII CO

N

«»• CM «»

.C •»" "T" •^» Of f fN CO

X

45

Table 3.6 Relative permeability of cold rolled 11 gauge steelto be used in construction of welded shieldedenclosures as a function of frequency.

Frequency (GHz) Relative Permeability

0.03 1000

0.05 1000

0.10 1000

0.15 350

0.20 100

0.30 30

0.50 8

0.70 4

1.00 2

1.50 1.5

2.00 1

> 2.00 1

46

aioi ut-i JQ Oi

' 3M OCQ u

DiUHJ

Pli M^ lu

Dm

fe^

a;

J=4J

1+-I

o

co

•1-1

AJ

CC

dr-(

«>0)

So

Q)

4-)

c ,

iH S-<D

T) JDo; Ew fO3 -Co

Co CD•H ciJ

rJ +Ju fOC S-

0) <uE JD3 S-'.-4 O)4-i >W OJ

c S-•1-1 *'S,^

s:d) LU

h-4J

-otw <u

O g

—

fO

F utn to

txOO

o^ CI-

o UJo <

1—1 LOOQ :3

.—

1

<r

OJ

i-i

3W)

fe

47

oP-.

<

CD

4-1

CO

d

CO

>CD

O-U)

nCD

md

4J

cCD

efi•H l-l

;3 <D

cr ^CD p:

CO

H Ocy:i

M bUZ C

•H4-1

CD CO

,r: U4-1 CD

HC+-I CD

o >(D

!-i

X. \.o^sCC3 W5-1 HWo CD

4-1 i-H

O CO

^ OP-i CO

CD

bO

48

2at r—T—I I I 1 1 M I I I I I 1 1 1

1

1—I—I I I I I II

or3:

la -

i— Maximum

=—' Averaqe

^- -Minimum

100 1000

5a( 1—I—I I II 1 1

1

1—T—I I I

Zn

CO

1—I I I U I I

i

:

MaximumAverageMinimum

- Maximum

^" Average

~ Minimum

1000Frequency (MHz)

Figure 4.3 Measured VS'JR of (a) top plate, (b) back plate, and (c) side

plate TEM line, from 1 to 2000 MHz, transmittina into the

1/10 scaled TEM/ reverberating chamber.

49

ae o)fC .c

4->

cuE o•r- -l->

r— E•r—

,*—>.

N CD:r: eCD -r-

+JO +J

• -r—— ECVJ

03

totrans

r^ O NNI c5

(J4-> OfT3

Nii/ .— 00 i.Cli— O)

D- O E>^ O +-> fO

-M JEu O Uc OJ .

J= — CD+-> ^— E

3 4- fO +->

O E fOu E S^

Vce: CU OJ3 -!-> -Q

L.C/0 E S->» fO CD

(\J li.>

CU E O)• -M S- S-

G)

Figure

4.4

Measured

composi

the

broadband

ho

1/10

scaled

TEM/

CVl

o(S)

dMSA

50

c: -Dfo c: <u

S- I—-—^ O 03cnsz uc to•r- -a+-> c o+-> ro t

—

•I- JD -^-^

E -o ,—CO roC O QJfO S- -CS- JD +->

+->—' O) E-CI •—

O) +->

c -—^•r- T3 N1— E rr

to CDO)

(O ^-- '

t— N 00D. in I—

enO- oC O -(-)

-MI— O

O)-c: o I—4-> +->

dJ oen

-(-> n •1

—

QJ CD cuJD c C_)

-r- cu>, > S-U •1

—

E <u -D .

OJ u E S-1

—

O) n3 O)o s- JD•r- •>— cn F^- E roM- fD •t— -EOJ E +-> U

E 4->

en a> •1— CDE -M F E

• 1

—

E U) •1

—

r ro E -MQ. (O fOr3 o S- S-o •r— +-> O)u T3 V -QO S-

LO CUro >E CUE S-CU ^^

CU O E LxJ

s: .— (o t—

T3 -i-

cu s-s- cu13 Q.(AfO C75

Ln

cu

51

o

s-o+->

o

0.01

0. 1-

Frequency (GHz)20

O)

o.Xa;

too

a;s-oOJ

+-)

o

4->

20 1 1 1 1 1 I i TT 1 1 1 1 T 1 "T

10K\»

-^4t

M

(b)

M

n

-

' — » '

«

1i , 1 1 I .J 1 .1 , i i 1 1 i

1 11 i 1

0.2Frequency (GHz)

20

Figure 4.6 (a) Theoretical and experimental Q of the USAEPG 1/10 scaled chamberand (b) ratio of theoretical to experimental Q 0.2 to 18 GHz

52

1—I I M I II 11

—

I I I I I II

10 I aeFrequency (MHz)

1 eeFreQuency (MHz)

50

» 30E

ZZ

1 1—I I II I 1

1

T 1 1 I M I I I 1 1 1 I I I II I

xxxx — X Componentyyyy ~ Y Component

40j— zzzz — Z ComponentTTTT - Tot«l Component

le

-10

(c) Average

-* « T T T T^a.

J I I Mini . I I I 1 ri I I I

50 .

10 loaFrequency (MHz)

~ 30EN>

2 20

• 10

-1—I

—

I I M I II 1—I—I I I I 1 1

1

1—I—I Mini Y'

Kxxx — X Componentyyyy — Y Component

— zzzz — 2 ComponentTTTT - Totel Component

10

(c) Maximum

_* 4- T T T Tt;

= :r—n-^r-r-r^r^^^^^^^yf\l ty-v.^;^

I 1 1 1 1

1

II

I

10 1 00

Frequency (MHz)

Figure 4.7 E-field components measured at the center of the 1/10 scaled chamber using a

NIST isotropic (5 cm dipole) probe (position r3 in figure 4.14) with chamber

excited by: (a) top plate, (b) back plate, and (c) side plate. Net input

power is 1 W.

53

r-^

OJ.

NI

X

c(13

rvj

LjJ

S)

ID

3 Qn CM

(^ /Aep) PLQLc^ -3

rn G

54

T—I—I—r

T—I—I—r

T—I—r

m—I—

r

k

J I I L i J 1 J L i J I I L i J I I L i J-^l I L

S

G

0) -

X. ^4J . .

--I c•U O }-4

rt 0^ oX

•o •

i-l )-( TJ0) o C•H K m<+-( N-' 42

-dtd (D cfl

P O6 cC >-l

3 .-1 ^e P-M-( -OX ^ ccC O tfl

E CO

-Q Of=^ s :d\^ -o

N > CO O—' tHI O ^-1

U CM • • (U

x^w od T-l

X (1)

•p )-l -^rt (U o^ ^ -^-^

c <u 6

G)C rt TJ<u 4:: C

-2 bO CO

r -

/-^

(U ^ ?, •

OJ r-lL •—

Ix.-O fC PL,

CD <-)

}_, to .

"-1 „ -P

d !-i

cr'— 0)

0) \ >J^

.— v_^

OJ 0)

}-i -*-; 4-1

S -^ CC

a "^

ij oj 20. ^ ^-v

(D(9in (VJ

O)u

ON

<1>

ubO•H

(Fi^ap) •'•d

55

oc

. E -X_lo

u

(S i«

^ Ql

< . 1

>UZ—

)

V h-

^^^^^aaS

CTE

•1

—

fO EO)

<X)

o>O)

Xs- E

UJ I.D

I— O-nOJ X

ro E

to CM

o O.-1\ wr-l

Oe> (U

Pli •-1

w a<C/3

P in1-1

OJ

4=U

oiwou

en CQ) 0}

rC Eo Q)

AJ O0) cC

^ .-<

« a

i-l bO« cc •Ho s•H oU XO w0) •

w uu 0)

0) xCQ XI »-l

M e oO n) CO

5-1 x: XIU o CO

0)

U

•I-l

56

XJ o^ c

3or

n

Cui/A HP) Pl»»J-3 CUJ/A HP) Pl»lJ-3

CM

O

I I I II

I I I II

I I I I 1 n I I

(•"XA aP) Pl»M-3

SM

I

00

(U

CD

(".'A aP) pi»»j-3

57

(ui^A aP) Pl»>J-3

(S Q S sw —

(ui/A BP) Pl»VJ-3

58

0} -o-C c•u cfl

c to•H

W •UJJ 4-)

C •r-l

0) EE wcu Cu CO

3 }-i

w iJ

/-v to

OJ (U

NI

E

5-J

ECO

00

iD0)

^^ Ou

CO &> c o .

c f^CO

c4-> ^^cC

&?^3 bO

c.H ;3

u •H ^ •?

V>

.1-1

i.0)

O 3 C

U-CU

5-1

•Hr-l

E X. CO

O 4->

>-i •H "Oiw

T3 (u

a> ^ w4J E CO

rt CO C1—1 x; cd O 0)

o 4-)

I-l CCO a CO

o OJr— bO

CO CO c-o O .H

LO >OJ •1-1

1— O (U

M- ^ CJ

LU .-1 5-1

(W/A HP) Pl»M-3

59

Frequency (GHz)

50

43 -

n•o

EU\

Eu

Frequency (GHz)

Figure 4.13 Measured tuner effectiveness in the 1/10 scaled chamber using- (a)receiving antenna power measurements (same receiving antennas as figure4.5), and (b) 1 cm dipole probe E- field measurements.

60

CD

oo

C-;C I-l

ca

-M Hfn 4J

u cfl

Q) CI-

.^ wS-fl) bD> Ca; 1-1

S-3

?~ wLU CO

h- a;

E-oO) u

1

—

oro 4-1

U(/) CQ

o ^I-l o\ S-i

I—

1

ao o(^ •Hw P-< OC/D S-I

^ 4J

O0) W^ •H4-1

otHO 4J

a>

^ ao 0)

u o(U cfl

^ I-l

w a4J

I-l bO >HCO c ec •r-t j-j

o s O•H o 14-I

JJ x; •Ho w c0) 3m

}-t •oQ) r-l

w ^ 0)

m E •r-l

o cfl tW>-i x: 1

CJ o w

<!r-l

<}

0)

S-l

:3

bO

O IS'

;» m "v o

NLaJ

rviB-l

("J/AaP) PL9L:^-3

X03

(LU/AaP) PLSLi-3

NXr

iX) m —T o)

<7.^1 1v ^ _^

9 n ^i «•

X

m ^ OB

n>- .riu a

(i^/Aap) PL3U-3 (lu/AaP) PL9Li-3

(U

en

62

IE .1» N

IS (<:*n

S-

>

I I I j I I I I

II 1 I I

I

I II

I I I I

to •" 10

m Xmu M in Xrvu nj in aw^ nj m CO

Cc

zeac

•3

s

IV

Zr

Xa

J c•3agL

I.

Sm

s

(w/ABP) PL9.L:^-3 liJ/AaP) PL9!-i-3

^ m Ty ^ —• m M <•

XUJ

O)en03S-

>

T m oOMD

Pl>- .rrv w

-I—I I I ' 'I L J I I I I I I I L.J I '- I

' ' ' I L_lI. I- I I I 1

'• pn ifcp' ' 1 1 . 1 1 1 1 1

—1—1—]—1—1—n—1—r- r—

n

^^— ^^^C^-^f ~

— ^^fft^mSl w —— ^K^w —-

J)rn''*^W"-

~» (ao.-«»

~

" '~

-jrtfjaplJT r^

-

-

J) •

.10 nn -

- '—i™'"!

' ^^j- y X :

T fii* -v; oV rx* -xj Xr cwc M . a>

— T CMC•r l<Me

NJ .-^ Xa; - o

"

~

1 1L \ t" J~

CU 1 II 1~ cr> « i»- -c X ——

2 ;^^ *u X —

— ^K- ^ s —— S- T ^~ ^ X —- cu » •- nj n -- ^ •— -y .-: o -

~ «• •- :\J o

- l»l— M w -

1 1 1 1

V la")- M

,.L 1 1 d t 1 1 1 1

m

A. 1

Jli

IS

(S ^NIz

B Q SV P) {\J

IS ss o in V

o sm rvj

Q S IS

(uj/AaP) PLaL.^-3 (UJ/AHP) PLaLi-3

CTi COS- C O)<D •.- 13JD 4-> I

—

E -M fti

rt3 -r- >U CO CD

C CD"O 03 (tS

I— +->

nDUCO CD

CD>

O 2 JD

I— CO X3CU C

CU I— 03-C O+-> CL E

•r- 13c -o E*"

E ><CO O 034-> Ec unCU-— --^c: ro

I/) -^

—

Q. OJ

E -Q ••

O O CU(_)£-+->

C2. 03-O r—•— U Cl<U T-•r- C- CLM- o o

1 s- +->

UJ +->

O EM- 00 Oo •.- s-

4-e 1—O C>0 s-•r— I—I OJM ^ s=3 O

r ^ -M-M +-) 13I/) -r— CL•-- S c-o -1-

"OI— CU +-'

03 E- CU•r- 13 E+-> CO03 03 :sD. CUC^ E I—

CUS-

cri

63

I I I I I

II I I I

II I I I

I

nj

S Is rPI w

>oc

a

(ui/AaP) PL9Li-3 (U1//QP) PL9L:^-3

T—I—n—I

—

III,

(w/Aap) PL9L.^-3

_ Q

_ M

_l I ! ' I -I I I I—Ip^

s r

G S cs s s1

in V n w

JD'

aOJM(OS-(Uc(Ucr,

s-0)jaEfOx:u-oO) •

I— Lr>

03 r—U •

O 0)r— S-

1— cn

+-> E

•1- <a•4->

(/) (O+J -ocCU EE OO S-

O T3O O)

ET3 T-^ EQJ S-•I- O)M- +->

I O)LU "O

-O "<U S-S- OJ=5 Sc/) O(O o.

E «->

Z5'4- Q.O E

3 +JE <U•r- EXfO 2

E x:ra +->

OJ SCDro CUS- -MOJ ro> I—m Q.

T3 Q-O) O+->

O <Uo x:E 4J

(ui/AaP) PL9Li-3

^

OJ

64

T-l

—

I I I I I Ij I I r

^ ifl -

N

Xre

s

' NIr

I I I ' I I ' ' I'

I ' ' 1 . . . . I . J-i^/h^P) PL9L:^-3

r^-I—I—r—1—I—[—1—I—I I

I

I—I I I

^«Mr.^"•jm-Tn/2^ ~"'

- u Br >:.

4->

O

X

-!_j I I I I I I I I _i ''(i^/AaP) PL9Li-3

s

I

T—I—I—r—1—I—r—1—I I I

•s fu^n CD

'i-^^ir-an- ^-r,

^- Id"*

X(O

I I ' I'

I I ' ' I ' I I I I' ' I

' ' ' ^ -» -^ '

(S S S) IS s s sin « n cvj — —

(lii/AaP) PL9.Li-3

1 f*

fc^j 1—1—TT r [ r-T—1 1 1

1—r—I—1

1 1 . 1,

. , 1 I

z~

- —-iSP^ -~

nj j[*>af n~

~ Cff "^^Jfc ~- -

"

"

-^IJSiJ^^^,'^*'"

~

__ ' r M ti^ n-«j rrr . 'U

«> *•«/•VST m^akjm- -i

-

m -Ji^ .

^£-^^^ :"

u"" '• ^---P-. % -— u M a

- — ^J u n . «D -

UJ

- .-vi ID n -«5

t \ J^^'cL wJR- X

fO ^ X B&^rS-s:

1 \ ^'^1— *" '-J n tf ^CD "— — % DIB «B -- — X rrr «B --• — ivj m'j^o -- — "v. nam -- - i\j <me« -- — -J nOBB -

- ~ -• -

—1—L. , , 1 , , . . 1 . , , . 1 , . .

s

I

cs

Q C3 Q Q S Q Qin V PI M — —

I

(Lu/Aap) pi^i^-i

65

T—I I II

I I—I—I I I I I I

--^ MVi

:«3»^

.-To n «)

-•=-_=i^ " 'TfJ?.^.^ -

\

(UenrOS-Qi>

Lr

I

' ' ' I

SS8

Dig,

•umar

•• vs

fC•4-J

O+J

<U

(OS-

>

s^ NX£

_I_J_ I I I I I I

(li^/Aap) PL9Li-3

sin

Qm

sN

(lu/Aap) PL9L1-

I

I I I I

I

I I r I j I 1 1 t

3 n.vi» ru

S-

>

J I I I I I I I I

-1—n—r^B^i . • '1

^1 1 1

^Hrfl^nj 3f

_ V ^^V^fam!_ —— -*3i(pf^v •"

~

C-3R"^-

"

~iw «tn.«- ui m

"

n"•^"' -*g PIT _ (0 Im

- BCD- »»• ^«^»»-""•a^S A

I_jl- ^n/'S/'

na1

- Ti-« ' &• -V rS- -1- — U) ru

- n. K. - nj ID <n «- •

- M .fi (n • at

- "» 4 1rjfi^_ I \I Sh -

UJ\ Wii

OJ ^ fVJ n^ ^fl> —^ « nj m;B «! -_ a !\J P»S «l -L. rO M rj nia«5 -_ S- — i\j P)»« -*

- a;> - vi ra-m -

- =3:- rg no*D -

- - (M • « -

"•

J- I-L. 1 . , , , 1 , . . , 1 . , , , 1 . . ,

-o OOJ S-S- «4-

to S-fO 0)CU 2E O

Q.i.d) 4->

JD Z5

E CLrO Ex: T-O-a CUO) c:

"to 3uto I

—

O C7)

I— EI— 4J

+->

<U -r-^ E+-> if)

EE fO•I- s-

+->

to+-> E toE O) CDO) -E 3E 5 .—o foCL--~. >E c/1

O O) CD

<a r— CT>

O rO

T3 Q- S-I— •.- QJOJ T3 >•I- rOM- EUJ JD

LDt|

O T3lO E

E CU fC

O -Q•.- O E=5 a. EjD -1-

o x.•r- n3CL. Eo

•1- S- '—

^

-O +-> fO

OI— to(T3 •- ••

•r- CU-t-> I— +->

tC C/) fOQ-i—I t—C/) ^ Q-

(uJ/AaP) PL9!-^-3 (ui/Aap) PL9!-i-3

S-+->

00

OJ

CD

66

I I I I I I

S X. s r

ct3

•L.

C"'/'A BP) Pl»»J-3

s >,Ci JD«4

(U -a<u4-)

03

O)CD

s-

OJJ2EfOx:

N o5 T8 r am O) •

1— r^X fO

—

u u •

c•

to t^d-

3a o O)

s r— S-

L --^ zs

L.

the

1

n

fiq

(W/'A BP) Pl»IJ-3

•r- (O+->

CO ro4J -a

OJ EE OO S-ClM-Eo -ou OJ

c-o •>-

•— EO) s-•r- <U4- -M

I O)UJ "C

-o »

cu s-S- O)=5 5CO Oro O.OJE +J

4- Q.O E

:3 +->

E OJ•t- EXfO 3

Ero

QJ ScnfO CUS- +->

OJ (O> 1

—

(O Q.

-a J:^

O) Ox: fO+-) J3oo O)

OO 4->

CO

cu

tW/'A BP) Pl»IJ-3 (XA BP) Pl«»J-3

67

s

I I I;

I I I II

I I I II

I I I I

N

X

cu —

I WIM -

''''«s s s

N

''''''

8 N- Ir

(lu/AeP) PL9L^-3

8IMI

I I I II

I

"S&£S^.I0VI ^^Xi=

•S> ••CB rt to*

- 1?W=

o+J

Xto

8..8 N-Ir

•'*'**'' I' '•" ''"^QSS88888 (Onvmiu— — cvi—

(Ul/Aap) PL9L:^-3 ' ' ^

II

I I I I 1 I I I II

» I I II

I I I II

I I I II I I I I

Jf^

n «e^'^R,

^£p

X

XITS

I

8I '*' I

8n 8

I I I I J I I I I

8 ^8 N- Ir

(lu/AaP) PL9!-:^-3

8 8— NI I

11 1

1

11

1

1

1

11 1 1

1

11

1

1

1

'

I' '

'

'I

' '

'

l*(«-^ IS

'^-^i'^-??.I V^—

UJ

X

8 8 8« «

. I .

8N

I I I I I lilt..

(Ul/Aap) PL9.Li-3

8 ^8 N- I

8NI

CUs-

68

'I

•'

•' I

I I I I I II

I I I II

I I I II

I I I r

"^ ''

^aflyrjn ' ^^•i™ --

jp -i 4^T "^

entos-

>w -

SHI

I I I I I

oj_Ls

'•'

S ^S N— I

r

I I I I I II ' ' ' I

*ssr-

re+JO+->

O)cr

&_

>

. . f *'''' '''

S N- I

j:

stuI

sin

S sN

sniI

Ciu/A BP) Pl»IJ-3 C«^A HP) Pt«tJ-3

+J -i-•r- </)

T3 OS- 4-

I/) s-

OJ sE o

Q.S-QJ +->

JO =5

E Q-fo c:x: •-u

+->

-a O)O) c:

m 3:aCO 1

—

o en

t— +->

<U -r-^ E+-> 10

c•r- S-

+->

CO4-> C C/J

c <D a»

II I I I

II I I I

II I I I

I

"«. "^

(Ucn

s-

>

1 I I I I I I I I I I t I I .... I ... t .... 1 .. .BOBSSSSSn M •• mt f^I I

s ^S N- Ir

uc

aac

I I I I flMLJ '[

I I I 'I

' I ' 1' ' ' '

I' '

' ' I '

^ IS

ft. n T

— \Ir.«-~ _«r>— IS "" »

CU

fOs-QJ>

8 ^S N- I

r

n

''<''*'*''*'« ^

B Q B S (9 s a« n , w -

BM

I i

CW/'A aP) P|»(j-3(ui^A BPJ Pl«»J-3

o foa.'-^ >E 00

OJ CUCJ .— CDO 03

T3 ex S-1— •.- O)01 -o >•I- 034- E

I U '-^UJ JD

LOI)

o -oto C

C d) 03O -Q•.- O E+-> S- 3Z5 Q- EJD •!-

•r- -- XS- -r- ro-M CL ECO O-O -1-J 03

OI— COfO -r- ••

•I- O)-4-J I— +->

03 00 03D-i—1 .—OO ^ Q.

CT^

O)

69

IH I I »

II ' I I

II I I I

II I I I

II I I 1

' I^J ' < • <

ss >,^ X)nj

oO)4->

n3S-OJc:

O)CD

S-<u

E03

14-C

s T us rm ^ o

X I— criu (C 1

—

c•3

u •

to «;a-

a-mL

o <uI— s-

I'-yA BP) Pi»ij-3(•y'A BP3 Pl»IJ-3

I— O)

•r- 03+J

(/) fO-MT3

O) Ec oo s-Q-M-Eo -ao cu

c-o -1-

I I II

I I I I

^ iV ^^ \ en\-

( "^V \ s-

e( >\<

i\1

i

X ^ VfO >

^X

LxJ

<

_ s

G Z

in

("'/'A BP) Pi8ij_3

CU S-•I- <U14- 4J

I OJLU T3

XI •>

O) S-S- OJ=5 SlO o03 D-O)E 4->

Z54- Q.O C

ZS 4->

E cu•r- CXfO 3

03 -l->

CU iro CUS- +->

CU 03> r-03 Q-

"O <UCU "OJZ -r-

-M CO

OO CU

E -eC>0 +->

oCM

CU

(•"•A BP) Pl»»J-3

70

toroC S-E OJCU JD4-> Ee fofO J^

in

n

)of

the

hor

1/10

scaled

^% to <Uto sz

N O -M

I ""cu

t3 E -a3 •!-

y^ E 00•r- CE -r-

Q >•1

—

E -a

n —' O)u

c U 1—

<U

3 o ency •1- E

M- -r-

M- >L cu -r-

Ix.

d

coupling

itting/rec

40

GHz.

If)OJ E 4->

S- 1/5

aj13 E Oto fT3 C\Jro S-OJ +-> +->

S •—' 03

OJ

nj

Qp C^nOcJ/uid) sso-j uuo^sXs

71

1 r 1—

r

1 1 r

uD-

U HCE inU) MD z+J 4*

fd

TS "0

(U t)

L. 1.

J 3M Pi

fd

Q)

Z S****

ID

S

Q^

power

measurements

er

of

1W.

U) E OI'J

ermined

fro

net

input

p

^% 0) fO

N -E

I S- +->

13 -Q SEy^ ^ N

Q > CD

n 0) Or- ^

c COo o(/I +J

3 o o0- 1— oo

V (T3

L. <L) E-E E

u.

ields

in

t

iving

ante

LO

OJ

Figure

4.22

Maximum

E-f

of

the

rece

Qru

(Ui/AQP) Pl»lJ-3

72

Xen cc csi q:: Qi

^ ^ ^ ^ ^o o o o oO) h- Lo n V-

[] X l> 4^ <>

oLO .

CO<u

+J

x>

i=5-e:

un

o«4-

S-

o IV

in

oo >,\^) o

c:CD13crO)S-u_

LOc\l

o

LO

+->

-Mro

EO+->

fOci.

os-Q.

aN

Es-o

+->

03

U

C_)

CO

OJS-

CD

o

(m>i/ap) uoL:^enu9:^:^v

73

Table 4.1 Spatial variations in E fields measured inside the 1/10 scaledTEM/ reverberating chamber over the test volume defined by probes

shown in figure 4.14.

Frequency, MHz E-Fields Variation, dB

0.01-200 +10

300 ± 8

500 t 6

700 + 4

1000 ±32000 + 2

74

Table 4.2 Summary of operation characteristics of the 1/10scaled TEM/reverberating chamber.

1. TEM Region (10 kHz to 70 MHz)

Planar Field Simulation

1 kW Input Power Yields Approximately250 V/m vertical polarization and 80 V/m horizontalpolari zation

Uncertainty < ± 2 to 9 dB

2. Transition Region (70 MHz to 300 MHz)

TEM Field Plus Limited Higher-Order Modes

1 kW Input Power Yields Approximately 250 to 1000 V/m

Polarization Variable

Uncertainty = ± 10 dB

3. Reverberating Region (300 MHz to 40 GHz or Higher")

Complex Field Simulation

10 W Input Power Yields Approximately 100 to 300 V/m

Uncertainty = ± 4 to 8 dB

75

(U

-o uyl V 3

1. /\ T3 CO

§ /\ r-H OM Dr-Io •— • \ •H OZ. ^ \, X. C

(J \ y MCden U ^\c i. \a \Q. __ \u

gj \U1 c

\I. Q.

c 5

? TS

5 ™L^ fli /^

X Y

—

uJC

,

o

ta /-1 1—1 4J

CO to

C 3/ C/ •H 0)» / 4-1 4-1

-— /"V^ -^ Q.4-1

o<:s; \ /

<0 \ /i- \ /

*• \/

u

^

J?

3T3

Q.

n

01

cu

I.

3

u

_

c

L(J

1r u

1

u4J

c0)

11 tr L—-I (1

3L Cl

Ji Q.(J T-u+*

JD LQ

t.

M

c," tl

u CO ^ ' )

- u ^M M N *»

30)

00- o ^o_J

a. ~ oc•l-l

MH

0) -

O =5j.'

I.

e

L<j

c

uc

•M

£E

t.

Ul

3

JQ

cS c

•U

tHO

CO•H4-i

CO

1-1

>0)

<4-i

)-l

X!0)

C/)

rH

X)(L)

CO

c -QF

•H «3•P x:nJ u4-1

C aiCU c6d +jS-i fO4-1 S-CO <uC JD•rH S-

<D>

<+4 <DS-

?^s LUCO t—^4

bO -acfl (U-H P^-o CO

to^

1

—

CQ >—

in

(1)

5-1

bO•H

76

1 Pc. 7.6 cm Absorber

(b)

1 Pc. 12.7 cm Absorber

(c)

1 Pc. 20.3 cm Absorber

(d)

2 Pc. 30.5 cm Absorber

(f)

4 Pc. 30.5 cm Absorber

(g)

8 Pc. 30.5 cm Absorber

h'

2.5

Time (ps)

5

Figure 5.2 Maximun vaijes of transmitted (positivo) ar.d received (neqstive) r-^

pulse waveforms in the 1/10 scaled TEM/reverberating chamber

determined with increasing amounts of rf absorber loading at 0.9 GHz

77

Time (ps) Time (ps)

Figure 5.3 Maximum values of transmitted and received rf pulse waveforms

inside empty (no absorber) 1/10 scaled TEM/reverberating chamber

determined at selected frequencies. Rf input pulse width is

5 ys, duty cycle is 0.001.

78

2.5Time (us)

Figure 5.4 Sheet 1 of 3

79

1.5

Tiine (ps)

Figure 5.4 Sheet 2 of 3.

80

Figure 5.4 Sheet 3 of 3

Figure 5.4 Maximum values of transmitted and received rf pulse waveforms inside

1/10 scaled TEM/ reverberating chamber determined with: (a) no absorber,

and (b) 1 piece of 20.3 cm X 61 cm X 61 cm (8" X 24" X 24") rf absorber

at selected frequencies.

81

-1 III 1 1 1 1—1—1

—

—Ill 1 1 1 1 T 1

nj

•

- Xo - u

o _J=^-

-^^^ :::^-'^- (U

_ .

<rc

i

-

J' 1 1 111 1 1 1 1 1 1 1

s s

(M) :»nd3no

tl I II I M I I1 III I I I I I

1 Is

(rt) ^nd^no

Mill 11 1 T llllll 1 1 1- mil 1 1 1 e

- _ c

-

NXOON

_

•

o ____.

/\

-y"

-

i

CEZ ^

^

<Mill 1 1 1

>_^^ IE

1 Mini 11.1 Mill 1 1 1 i

(M) and^no

(H) 3nd:ino

llllll I—I1 II I M I I—

r

nil I I I1 MM I I I I

1 MM I I I I1 1—Is

III I I I I I II I III! I I I II l—ls

(tt) andano (M) ind^no (rt) ind^no

82

n—I—Hill J I I—I

Mill ITn-(T-r-i—

r

T II 1 1 I I I—I—

r

CD

a.

_ I

T—1—I1 r I I I I—I r

_ o H

L

(M) and3no

III I I I I I 1 I I I I I I I—1

T"

(rt) andjno

iiiiiii I IImill 1 1 I—HUM 1 1 i

;

OC\J

+->

OJ<U

oo

IT)

in

s_

u_

(rt) ind^no

83

(W) ^nd^no

(rt) 3nd3no

. Hi

(rt) and:ino

CO

mil 1 1 1 T Mill 1 1 1 T mm 1rrk

JOL

•. A

1 Jz ^—^_l /

<=">" 5

5

1

N -

ac m

oI

I o> u ^ iT3

S 00 g _

l'

3

i

'k V^;—

—

=

11 11 1 1 1

—limi 1 1 1

1 tiiiii A,=

mil I I I—n III I I I

—

iiiiii I I I—mill i 1 1 1 ; lu

33O

I °

mil I 1 I iij111 1 1 I I I nil M 1

CO

^N 4J

a. (U

(U

^(U CO

•HH

)-l

•H

(rt) 3nd3no (rt) tind^no

""I ' iiiiiii I I'^ffSi I i-4f

mH

(rt) ^nd^no84

(M) 3nd5no

(n) 3nd:ino

T-r-T rrrm i i Mill 1 n T mil 1 1 1 in

- ^ ^ - t

rn

N

-1-, O '

i- OX

-s vO

uo

lU

<

ITU>

,

1

\^ SsV .

J_L 1 UJJJ 1 1 1 II 1 1 1' "1 * ra

(M) ^nd^no

III 1 1 iiimrr1—lllllll 1 1 'lllllll+-t IIIIIIH-I>—

.

\

:S

: ..-^^

o ^^•o y^vO r^ J*^,-^ cii ^^

-t r-> ^/<

X j^c f

r X>^

OU.l -UU

\

III 1 —lllllll 1 -1

lu 0)

B

X.or oi

C 00

II II I I I l_

^(n) :indano

e ^ ,—>'

H ^ S_H LU O)1— X3

S-T3 odJ (/)

1

—

J2(O rtJ

Uto Oo ^—[~

>>' +->

CL tocu E ^j:i O) +J4-> -a

S- -1-

OJ O) s"O -Q•1— E Qito fO toE -C 1

—

(rt) ^ndino

•r- <_) 13

^ S- x:o O T3

•4- M- ro EO CU O fO

> S-"^ ro CI. to

2 CL cu+J fO •>-

Ol d) oO) to TD Ejc 1

—

OJ O)t>0 13 E 13

Cl =3 a-

LO M- 1 S-. S- O) ^-

tn T3D O -QO) O) E O)S- > <->

:3 • 1— >> uC75 0) ^ <u

•1— u r^Ll_ <D -O <D

S- CU COE

<4- •r- Oo E -—

to O) H->

O) -M fO:3 O)

1

—

-U EfO QJ>

CD roOJ J3 4->

en Ero fO CDS- x: s-OJ U O)> Sre en

E too •1- +->

e 4-> EfO

^ Et OJ o)13 J2 s-

F S- 13•1

—

O) 1/5

X > fO(T3 OJ q;s: s- s:

in

Ln

O)S-

(rt) :ind3no (ft) and^no

85

uI/) I

—

•I- <£)s- Xo-M E

os-<U I—

Eto XJZo E

uenc: ro

+J os-OJ 4-jr> os-<D O)> O<U (Us- •--^ Q-

T3 J3O)—

-

fO "OU Eto fO

O •>

1— s-^ O)I— ja

s-OJ o.c to+J -Q

O+-> o

OJ-^+J+-> ••

•I- aiE Eto •!-

E to

S-4-> O)

XJI— 13fC -l-J

CDr

fOM-S- <u

-MS- fOo -MM- to

-D >,O) "O •

S- 03 S-•1

—

O) <uZ3 +-> J2CT to S-O) os- <^- to

o JDO) mE ^•r— CO ^-1— to s-

<£>

LO

(Us_

oi

(srt) aiiiLi aSueiio

86

1 I r

"3.

J 1 L ' I L J I I

OJ

OJ o

oO)13a-CD

OJ

OJ ID in Q G

o+->

Xs-O) E-Q <_>

Ero nxz •

o oC\J

enC 14-

•- o)->

03 OJS- U<U CDJD •-S- Q-OJ> I—OJS- -~-

I— -aE

a;

fO s_U OJ(/) J3

o o1— (/)

1— fO

<u o

O ra-M —

'

CD-o EQJ •1

—

+-> (/)

+-> 3•^E cuto -o

S- -r-

rofOCCD OJ•I- -Mto

^-

oM-

T3O)

cu

<T3 S-+-> Oin (/)

I ^>) toT3n3 ^-Ol s-4->

I/) E<4-

O<JD

O) os- en

O) OJto

I— s- vo

un

s-

(srt) aLUL:^ a5jeL|3

87

COa

oi.s-

oC_3

•

\'^\

0.?) GH z

(bK1

e

0.1

(J

s-S-

o«_3

0.1

0.5 1 2

Pulse Width (ys)

^\,^•v^

1.3

1^

C5Hz

(b)^X-

1

^^^a)

ca

uS-s_

o \ c...)

1

3H2

i —.»

1

________j_^

^^^ (a)

B

0.1

£2

oS-S-

o

Puis;e width (vs)

s

^ 4. z G^^z

i

( d)"^(a)

1

0.5 1

Pulse Width (ys^

0.1 0.5

Pulse Width

1

'us)

C3-a

o•->

oOJS-

oo

v

N 2. GH z

\ \

\^

(J)

(b) \

1

1

^^1

0.1 0.5 1

Pulse Width (ys)

uOJs_S-o

e

0.1

\

\.^\ 5. 65 C

1

;h^

ib:^

1 JT^-^0.5 1

Pulse Width (ys!

Figure 5.8 Sheet 1 of 2.

88

CO

U- 3

(J e<us_i-

oo1

V,\^N

8. 9 GH2

(t.-Vk - (a)

0.1 0.5 1

Pulse Width (us)

9

oofO

o \0) 'S-

o \\ 12! GH z

B

(b

\\\. (a )

rc .1 5 1 2 4

Pulse Width (ys)

9

cn

4\S-

o+->

<_) \o

\

\,16

1

GHz

0)s-

o \

\\

\Jb) \N a)

B\l ^^,

.^

C).l 0.5 1

Pulse Width (ys)2 4

^^ 9

o4-)

u(0U-

3

1?i GF\z

o

oCD z\s-

oo1

N

\(

(b)

"s\\e -— ^^

^^__C .1

Pu lS€

5

Jidtl1 (liS)

2 4

Figure 5.8 Sheet 2 of 2

Figure 5.8 Estimated correction factors for amplitude response of rftest pulses in the 1/10 scaled TEM/reverberating chamberat selected frequencies as a function of input pulse widthusing (a) no absorber, and (b) 1 piece of 20.3 cm x 61 cmX 61 cm rf absorber.

89

-a

u<us-i.

oo

CO•a

oa;i.

o

5

4

<

0.1 yS »(a)

^

2

1

0.5 vS

* —

2 yS «—

k

——

—

B .B 1- 3 2.9 c .6 • B 12 :e

Frequency (GHz'

A

(b)

3 ^^^^--^^^^ 0.1 ,JS «

-^^—

^

««t -^

2

~~-^---.^

«t

'^^-•

1

^

«

*

1 vs

— « 0.5 vs

,*

4

«

«

B "—-—*——

_

_ _ —

«

.8 1

.

3 2.S 5 . E B 12 le

Frequency (GHz)

Figure 5,9 Estimated correction factors for amplitude response of

rf test pulses in the 1/10 scaled TEM/reverberatingchamber at selected input pulse widths as a functionof frequency using (a) no absorber, and (b) 1 peice

of 20.3 cm X 61 cm x 61 cm rf absorber.

90

Table 5.1 Summary of influences of rf absorber loading on loss,

quality factor, and charge/decay time of the 1/10scaled TEM/reverberating chamber at 0.9 GHz.

AbsorberLoading

Min. ChamberLoss, dB

Max. Oual

Factority Measured

Charge/DecayTime, yS

None 7.0 10 418 0.29

1 pc. 7.6 cm 10.8 4 333 0.22

1 pc. 12.7 cm 11.1 % 043 0.19

1 pc. 20.3 cm 11.7 3 522 0.18

1 pc. 30.5 cm 12.2 3 139 0.14

2 pc. 30.5 cm 15.6 1 435 0.10

4 pc. 30.5 cm 15.9 1 339 0.10

8 pc. 30.5 cm 21.4 377 0.10

*Charge/decay time is measured from figure 5.2, based upon 63% riseto steady-state field amplitude inside chamber after excitationsource is turned on, or 63% decay from steady-state amplitudeafter source is turned off.

91

Table 5.2 Comparison of measured and calculated charge/decay timeof the 1/10 scaled TEM/reverberating chamber, determinedfrom chamber average loss measurements (figure 4.5) andfrom data shown in figure 5.5.

FrequencyGHz

Average ChamberLoss, dB

AverageQualityFactorX(10)4

CalculatedCharge/Decay

Time*ys

MeasuredCharge/Decay

Timeys

0.9 18 0.083 0.29 0.56

1.3 22 0.099 0.24 0.68

2.0 26 0.144 0.23 0.60

2.9 30.5 0.155 0.17 0.56

4.2 35 0.167 0.13 0.36

5.65 37.5 0.229 0.13 0.37

8.9 42 0.318 0.11 0.52

12.0 45 0.391 0.10 0.45

16.0 47 0.584 0.12 0.43

18.0 48 0.661 0.12 0.30

*Calculated charge/decay time is from 2Q/c

92

70.

G0 -

50

nO 40

i 30

20

10

-

1 1—I I I I I I I

1

1 I I I I I I I

1 1—I I I I I I I

Steel

I tI I I I I I I I 1 I I I I I II

0. 02 0.2 2

Frequency (GHz)20

Figure 6.1 Projected coupling efficiency (minimum loss) of 13 . 1 m x 24 . 1 m x

m TEM/reverberating chamber.

93

55

E\>m•o

iZI

u

E\>n

•o

I

u

50

45

40

350.02

55

50

45

40

35

0.2 2Frequancy (GHz)

1 1—I I I I I I I 1 \—I I I I I I I

(b)

1 1—I I I I 1 1

1

Zinc (RRDC RC

)

J I ' I I I I M J I I I I I I I II I I I I I I I I

0. 02 200.2 2Frequsnoy (GHz)

Figure 6.2 Projected E-fields in 13.1 m X 24.1 m X 38.7 m TEM/reverberating chamber

for 200 W input power determined from (a) projected chamber losses and

Q factors, and (b) measurements made in RADC chamber and NSWC chamber.

94

50

cu

S-

o

10

1

•

T I r 1 1

1 I I I I I I 1 1—I I I I I

J I I I I I I 1 I

0. 02J I I I I I I I I I I I I I I I I

0.2 2Frequency (GHz)

20

Figure 6.3 Projected charge time for 13.1 m X 24.1 m X 38.7 m TEM/

reverberating chamber.

95

g<CCLUQ.

OLJ_

O

DCLU

<oo

o <o gcc cc

CO ^ ^I UJ SW N ^2: s ^

Z h. NO 06

^O CO

>- X< E

g

I

N

o

LU

CO

Q_lLU

LJ.

GC<Z<Q-

oDCQ.

<COa-ILU

>•

I-

Q.

5

OQ.

QCOI£

>o10

QQo

I

Q.^

£

>oo

<DCLUOZ3

£

> mo "^

10 oCVJ T-

co

N Q

oCO

I

o

QLU

N 2

CO

I

omXoDCQ.Q.

<COQ

+)

II

>•

Z<I-DCLUO

> _

ZgCOz<cc

Q.

O-JLU

LL

LUI-

^

g<NDC<_JO0.

N3: zO go h-'^ <

• ^N S

CO UJ

£

>oo en04 -a

OzI-<DCLUCDDCLU>LUDC

DC

<LU

XLU

Oo

XoDCQ.Q.<COa-J

LU>

00

I

+1

>I-z<

> QJ

O Oo zeg 3

CO*-

LU"^

-J CO

<

cv CO

96

J^^oocoe-

Oo

CM

2

aivid aais

OE-

-J

M O

Q,

a:KWWE^jOCLCLS<

E-O

So:DCdQi 03

E^>:o-Ju<0,2C/] <

Pi

QUE-

bOC•H•CJ

i-H

0) •

•^^ U^ Q)

w6

t3 CO

C rCtfl o

>^ W)•u c•H •Hc U:3 cd

1-1

g (U

•rH J3v^

-o QJ

(U >•u Q)

cfl U•r-l \T3 sCO w>-l Hs bOw C

•iH

^ cn

o 13

C0) S;3 (U

cr •p

0) en

)-i >^M-i en

iJ 4->

CX C(U OJ

S EM 0)

)-i

D4-1 wO CO

QJ

E!=

CO U)

>-i U)

bO Q)

cC c•H 0)

73 >•Hu

^ oO 0)

O iwt-H 4-1

CQ 0)

a>

ubO

97

SWEEP OSCILLATOR

DUAL

CHANNEL

XY

RECORDER

WITH

TIME

BASEJ__

RF AMPLIFIER

1

DIRECTIONAL COUPLER

Q_Ci

oTQ

a.

c•H

I—

I

<u

w

CCO

wCo

w }-i

w (U

•H ^e S<u to

x:T) o0)

4J bOto C•H •H-o 4-)

Id to

1-1 U0)

S ,Qw »-l

(U

>^ >o a>

c )-i

0) \3 acr w0) H>-i

iw 00C

•p •rH

P- w0) dS

fU

IW 4-)

o to

>-.

tn

eto MU WbO (U

CO C•r-( (U

-o >

4-)

^ oO <U

O '4-1

i-l IH

CM

(U

bO

98

10 dB/div

MEAS UNCAL

u IfWfl^^^A^U I li I^ rfifvffpWf%\A¥mfijidiAlfi3 . 02

.

:

24 Gl-

30 dE Imi/i

r M 1'\^^

' Ll . J /r*^ if|ik4 4/Wi/l^/ftU^hwVf/%'^i^kA 1 V/Mi/iJfi4^^^w/TT^ -^\^^

)

CENTER 3.00 GHZRES BW 1 MHz VBW 300 kHz

SPAN 2 . 00 GHzSWP 20 . ms

Figure 7.3 Measured SE data of 3 cm X 6 cm TEM cell with 15 mm circular aperture

using the 1/10 scaled TEM/ reverberating chamber and swept frequency

measurement system.

99

0)

ja4J

bOc•Hm3

g0)4J<Q

>^W

C4J oc •H0) 4->

6 CO

<u 5-1

5-1 <U

:3 CXw otfl

0) <wE o

^ OJ

J-) -0•1-) oi-H e•i-l

^ s•iH w4.) Ha0) Q)

o x;w U3CO c

•HT)0) 5-1

p 0)

CO ^E eo CO

4J rC13 oCO

W)M-l cO •H

4J

e- CO

cO 5-<

M <U

tiO^CO 5-1

•H (D

•O >0)

^ 5-1

o \o s

I-l wCQ H

5-1

bO•H

100

e a>

<u ^u •uVI

>>tn •T3

Cp cfl

CQ)

E C0) oS-i •r-l

3 4J

w WtC •Hd) me C

CO

>^ M•u •u•Hr-l

•H 0)

rO rC•H 4->

+J

a(U Co •rH

tn

3m

TJ ^<D E4-) CO

CO ^ ^

•H O <rT) nCfl

>-l WJ 4->

c fnX) •H <;_

0) P 014J CO ntC v-l ob 0)

o 43 4-i-i U n3 Q)

CO > 00Q) 01

IM U T-JO \ Os FB UJCO H mS-i eUJCO Q) 4->

•H rC mT) 4->

(D

,^ bO s-O C O)o •H >I-t VI OJCQ S S-

in

r--

a>

u3CkO

101

COT3

.Q.+->

o

oQ.

70

60

50

40 -

30

T T 1—I I I I T T 1—I

—

r-TT

rrrr 1/10 scaled TEM/reverberating chamber

aaaa NIST anechoic chamber

nnnn NIST reverberating chamber

0.2J I

I I I I

Frequency (GHz)

'11120

Figure 7.6 Maximum response of NIST 1 cm dipole probe to an E-field of37 dBV/m generated in (a) the 1/10 scaled TEM/reverberatingchamber, (b) NIST 2.74 m X 3.05 m X 4.57 m reverberatingchamber, and (c) NIST 4.9 m X 6.7 m X 8.5 m anechoic chamber,

102

30

CO-a

(->

Q.

zsofOccO)+->

o

20 -

10 -

-IG

-

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

-/.r***'"''^'''^'--. -

-

-

-

aaaa NIST anechoic chamber '\it

rrrr 1/10 scaled TEM/reverberating chamber ^l

-

1 1 1 1 1 1 1 1 1 ! I 1 1 1 1 1_I_

-

0.2Frequency (GHz)

20

Figure 7.7 Maximiom powers received by a broadband (1 to 18 GHz) horn antenna in

an E-fieid of 37 dBV/m using: (a) 1/10 scal ed TEM/reverberating

chamber, and (b) NIST anechoic chamber.

103

30

20

dB10 -

-in0.2

1 1 1 11 I I I I

++++ Rnechoic Ch Minus Gainn¥rnii EPG TEM/rev chamber

1—I—I I I

J I I I I I I _L J I I I I

Frequency (GHz)20

Figure 7.8 Comparison of broadband horn antenna response in 1/10 scaled

TEM/reverberating chamber with its response minus free-space gammeasured in the NIST anechoic chamber.

CO

(/)

0)cd)>

o

CD

-a

(/I

80 1 1 1 1 1 1 1 1 1 1 1 1

70 - -

G0 - ^^^^W^\^|^^

-

50 - »^\£^^^ -

40ttttrrrr

Theoretical value ^^'^^^Measured in RADC chamber ^^^ >)f

"

30++++

In 1/10 scaled chamber with mode tune ^^

In 1/10 scaled chamber with mode stir•^ -

20 - -

101 1 1 1 1 1 i 1 1 1 1 . 1

0. 1

Figure 7.9

Frequency (GHz)

Comparison of SE of 3 cm X 6 cm TEM cell with 15 mm circular

aperture: (a) theoretical calculations, (b) measured in RADC

chamber, (c) measured in 1/10 scaled chamber with mode-tuned

approach, and (d) measured in 1/10 scaled chamber with mode-

stirred approach.104

BL-114A

(5-90)

U.S. DEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY

BIBLIOGRAPHIC DATASHEET

PUBUCATION OR REPORT NUMBER

NIST/TN-13422. PERFORMINQ ORQANIZATION REPORT NUMBER

PUBUCATION DATE

July 19904. TITLE AND SUBTITLE

Measurement and Evaluation of a TEM/Reverberating Chamber

5. AUTHOR(S)

Myron L. Crawford, Mark T. Ma, John M. Ladbury, and Bill F, Riddle

6. PERFORMINQ ORQANIZATION (IF JOINT OR OTHER THAN NIST, SEE INSTRUCTIONS)

U.S. DEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGYBOULDER, COLORADO 60303-3328

7. CONTRACT/QRANT NUMBER

8. TYPE OF REPORT AND PERIOD COVERED

9. SPONSORINQ ORQANIZATION NAME AND COMPLETE ADDRESS (STREET, CITY, STATE, ZIP)

Electromagnetic Environmental Effects DivisionU.S. Army Electronic Proving GroundFort Huachuca, Arizona 85613-7110

10. SUPPLEMENTARY NOTES

11. ABSTRACT (A 200-WORO OR LESS FACTUAL SUMMARY OF MOST SIQNIFICANT INFORMATION. IF DOCUMENT INCLUDES A SIGNIFICANT BIBUOQRAPHY ORUTERATURE SURVEY, MENTION IT HERE.)

This report summarizes the measurement and evaluation of a 1/10 scaled model TEM/reverberat-ing chamber developed as a single, integrated facility for testing radiated electromagneticcompatibility/vulnerability (EMC/V) of large systems over the frequency range, 10 kHz to

40 GHz. The facility consists of a large shielded enclosure configured as a transverseelectromagnetic (TEM), transmission line-driven, reverberating chamber. TEM mode testfields are generated at frequencies below multimode cutoff, and mode-stirred test fieldsare generated at frequencies above multimode cutoff. Both the chamber's cw and pulsedrf characteristics are measured and analyzed. The report also discusses the basis forsuch a development including the theoretical concepts, the advantages and limitations,the experimental approach for evaluating the operational parameters, and the proceduresfor using the chamber to perform EMC/V measurements. A full-scale chamber that will providea test volume of 8 m x 16 m x 30 m is proposed. Some projections of the full-scalechamber's estimated characteristics and operational parameters are also given.

12. KEY WORDS (6 TO 12 ENTRIES; ALPHABETICAL ORDER; CAPITAUZE ONLY PROPER NAMES; AND SEPARATE KEY WORDS BY SEMICOLONS)

cw and pulsed rf testing; radiated EM compatibility and vulnerability measurements;reverberating chamber; TEM cell

13. AVAILABIUTY

UNUMITED

FOR OFFICIAL DISTRIBUTION. DO NOT RELEASE TO NATIONAL TECHNICAL INFORMATION SERVICE (NTIS).

XORDER FROM SUPERINTENDENT OF DOCUMENTS, U.S. GOVERNMENT PRINTING OFFICE,WASHINGTON, DC 20402.

ORDER FROM NATIONAL TECHNICAL INFORMATION SERVICE (NTIS), SPRINGFIELD,VA 22161.

14. NUMBER OF PRINTED PAGES

116IS. PRICE

ELECTRONIC FORM

NIST.Technical Publications

Periodical

Journal of Research of the National Institute of Standards and Technology—Reports NIST research

and development in those disciplines of the physical and engineering sciences in which the Institute

is active. These include physics, chemistry, engineering, mathematics, and computer sciences.

Papers cover a broad range of subjects, with major emphasis on measurement methodology andthe basic technology underlying standardization. Also included from time to time are survey articles

on topics closely related to the Institute's technical and scientific programs. Issued six times a year.

Nonperiodicals

Monographs—Major contributions to the technical literature on various subjects related to the

Institute's scientific and technical activities.

Handbooks—Recommended codes of engineering and industrial practice (including safety codes) de-

veloped in cooperation with interested industries, professional organizations, and regulatory bodies.

Special Publications—Include proceedings of conferences sponsored by NIST, NIST annual reports,

and other special publications appropriate to this grouping such as wall charts, pocket cards, andbibliographies.

Applied Mathematics Series—Mathematical tables, manuals, and studies of special interest to physi-

cists, engineers, chemists, biologists, mathematicians, computer programmers, and others engaged in

'

scientific and technical work.

National Standard Reference Data Series—Provides quantitative data on the physical and chemicalproperties of materials, compiled from the world's literature and critically evaluated. Developed un-

der a worldwide program coordinated by NIST under the authority of the National Standard DataAct (Public Law 90-396). NOTE: The Journal of Physical and Chemical Reference Data (JPCRD)is published quarterly for NIST by the American Chemical Society (ACS) and the American Insti-

tute of Physics (AIP). Subscriptions, reprints, and supplements are available from ACS, 1155 Six-

teenth St., NW., Washington, DC 20056.

Building Science Series—Disseminates technical information developed at the Institute on building

materials, components, systems, and whole structures. The series presents research results, test

methods, and performance criteria related to the structural and environmental functions and the

durability and safety characteristics of building elements and systems.

Technical Notes—Studies or reports which are complete in themselves but restrictive in their treat-

ment of a subject. Analogous to monographs but not so comprehensive in scope or definitive in

treatment of the subject area. Often serve as a vehicle for final reports of work performed at NISTunder the sponsorship of other government agencies.

Voluntary Product Standards—Developed under procedures published by the Department of Com-merce in Part 10, Title 15, of the Code of Federal Regulations. The standards establish nationally

recognized requirements for products, and provide all concerned interests with a basis for commonunderstanding of the characteristics of the products. NIST administers this program as a supplementto the activities of the private sector standardizing organizations.

Consumer Information Series—Practical information, based on NIST research and experience, cov-

ering areas of interest to the consumer. Easily understandable language and illustrations provide use-

ful background knowledge for shopping in today's technological marketplace.

Order the above NIST publications from: Superintendent of Documents, Government Printing Office,

Washington, DC 20402.

Order the following NIST publications—FIPS and NISTIRs—from the National Technical Information

Service, Springfield, VA 22161.

Federal Information Processing Standards Publications (FIPS PUB)—Publications in this series col-

lectively constitute the Federal Information Processing Standards Register. The Register serves as

the official source of information in the Federal Government regarding standards issued by NISTpursuant to the Federal Property and Administrative Services Act of 1949 as amended. Public Law89-306 (79 Stat. 1127), and as implemented by Executive Order 11717 (38 FR 12315, dated May 11,

1973) and Part 6 of Title 15 CFR (Code of Federal Regulations).

NIST Interagency Reports (NISTIR)—A special series of interim or final reports on work performed

by NIST for outside sponsors (both government and non-government). In general, initial distribu-

tion is handled by the sponsor; public distribution is by the National Technical Information Service,

Springfield, VA 22161, in paper copy or microfiche form.

U.S. DEPARTMENT OF COMMERCENational Institute of Standards and Technology

(formerly National Bureau of Standards)

325 BroadwayBoulder. Colorado 80303-3328

OFFICIAL BUSINESSPENALTY FOR PRIVATE USE. $300