High Field Strength in a Large Volume:
The Intrinsic Reverberation Chamber
Frank B.J. Leferink
HollandseSignaalapparaten
B.V.
EnvironmentalTest Laboratory
P.O. Box 42,755O GD Hengelo,The Netherlands
leferink@signaal.nl
An intrinsic reverberationchamberwith non-parallel
walls, ceiling not parallel to the floor, at most two walls placed
perpendicularand with fixed field diffusersis presented.Inside
this room no eigenmodesexist and a diffuse, statistically
uniform field is createdwithout the use of a mechanicalmode
stirrer.As a result,test time can be reduceddrasticallycompared
to modestirredreverberationchambers.
Abstract:
INTRODUCTION
One of the main issuesof EM1 testingis the establishmentof
a high field strengthin a large volume at reasonablecost. The
creation of a field strengthof more than 6OOV/m(HIRF) is
extremelydifficult and costly. The reverberationchamberis one
of the possible test techniquescapable of creating high field
strengths.
The reverberation chamber generally consists of a
rectangulartest room with metal walls and a stirrer, usually in
the form of a large paddle,nearthe ceiling of the chamber.The
equipment under test (EUT) is placed in the chamber and
exposedto an electromagneticfield during which time the stirrer
slowly revolves.The averageresponseof the EUT to the field is
found by integratingthe responseover the time period cf one
revolution of the stirrer. The metal walls of the cavity allow a
large field to be built up inside the chamber.At the sametime,
the stirrer smoothsout the sharpnulls of the field; thesenulls are
usually present in such a resonant structure. The EUT is
therefore exposed to a high field level consisting of several
differentpolarisations[NBS Note 1092,Crawford,Hatfield].
REVERBERATION ROOM BASICS
When electromagneticenergy is fed into a (rectangular)
chambera numberof different chamberresonances,sometimes
referredto as normal modesor eigenmodes,are activated.The
type of mode excited dependson how the initial wave is
reflectedand returnsto the point of excitationin the samephase
and direction as the initial wave. In a rectangularchamberthe
simplesteigenmodeis the axial mode in which the component
wavestravels along one axis, one dimensional,parallel to two
wall pairs as shown in Figure 1. The power density in the
chambervariesas shownby the curve.
Figure 1: Electric field pattern, axial mode (4,0,0)
Another type of eigenmodeis one in which the component
waves are parallel to one pair of walls but are oblique to the
other two pairs, two dimensional,and is termed a tangential
mode.Figure2 showsthis type of eigenmode
I’
/. .. ..-.
2”
,, :
/
.........
...
,’
................
...........
i
Figure 2: Electric field pattern. tatlgelltial mode (2,l.O)
The third type of an eigenmode is one in which the
componentwavesare parallel to noneof the threewall pairs and
is termedobliquemode.as shown in Figure3.
,. ‘,....
,, ,,., ,,
.,
1’
,I
/
/
/”
C------~.~-......._.,,__,.._,,__,,_,,(
_,,
‘I ’
“.-” ..” ...^..~......,,..,,____
j.’
Figure 3: Electrfc j?eld patter-u. oblique mt)de (2.1.I)
The resonantfrequencyf of eacheigenmodein a rectangular
chambercan be calculatedfrom
f 2d(yy +(fy +(?)Z
W
where f
c
1
L
ndwb
MHz1
(1)
frequency[Hz]
speedof electromagnetic
wave[m/s]
lengthof chamber[m]
widthof chamber[m]
heigthof chamber[m]
integernumbers
To evaluate the frequencies for axial modes, two of the
integers are set to zero while for the tangential and oblique
modes one of the integers and none of the integers are set to
zero respectively.It can be shown that the axial and tangential
modes dominate at low frequencieswhile the oblique modes
dominateat high frequencies.
The total number of chamber resonancesN occuring in a
rectangularchamber in the frequencyrange 0 tof is [Briiel &
Kjzr]:
N- 47CV.f3+-nS.f2+k.f
3
where V
s
L
c3
4
c2
2
ANALOGY
ACOUSTICS
AND ELECTROMAGNETICS
The reverberation chamber modal theory as presented is
completely equal to that used in acoustics by replacing the
words
l
room with chamber(more common in electromagnetics)
l
diffuser with mode stirrer.
Both worlds are similar, despite the vector nature of
electromagnetic waves compared to scalar acoustic waves.
Reverberation rooms for acoustics were investigated by
Sabine(1922), Knudsen(1932) and Briiel(l951) amongothers.
THE INTRINSIC
REVERBERATION
CHAMBER
The responseof an EUT to the field inside the conventional
reverberationchamberis found by integratingthe responseover
the time period of one revolution of the stirrer. In a large
reverberationchambera large stirrer has to be applied in order
to guaranteea sufficiently uniform field. The rotation time of
this large and slow stirrer will determine the test time
completely.
A rectangular reverberation chamber without and with a
stirrer is drawn in Figure 4 and 5 respectively. A tangential
mode hasbeendrawn in thesefigures.
c
volume[m’]
surface[m’]
roomedges[ml
In statisticaltreatmentof electromagneticsit is assumedthat
the power densityis diffuse, i.e.:
l
the energy density in the chamberis uniform everywhere,
l
the energy flow in all directions is the same,
l
the polarisation betweenall waves is random and
l
the phasebetweenall wavesis random.
These requirementsare fulfilled if there are a large number of
eigenmodesin the room, which is the case above a certain
frequency[NBS Note 1092,Crawford, Hatfield].
In order to improve the uniformity a mode stirrer is
commonlyappliedfor the following reasons:
l
the stirrer reducesthe spatial varianceof the field strength
in the chamberwhich improves the accuracyof estimates
of the space-averagedfield strength;
l
the rotating stirrer producesa modulation of the power
tlow from the sourceinto the chamberwhich usually
makesthe field strengthof the source(at low frequencies)
somewhatless dependentof sourceposition in the
reverberationchamber.
The effectivenessof mode stirrers dependsprimarily on their
size. The stirrer should, therefore, be as large as the chamber
dimensionspermit [Briiel & Kjxr, IS0 354, Wu].
Figure 4: Tangential mode in reverberation room wirhout mode
stirrer
Figure 5: Tangential mode iI1 reveri~emtion rot~m with mode
stirrer
Instead of using a rectangularchamber with a mechanical
mode stirrer a chamberwith
l
no walls parallel,
l
at most one wall placed perpendicularto anotherwall,
l
dimensionsnot being a multiple of each other (i.e. 1=3,
w=4 and h=S) and
l
fitted with curved, fixed diffusers with a surfaceareaof at
least25% of the total chamber area.
resultsinstantaneously in a uniformly distributedfield. Note: for
wall, read wall, ceiling or floor. As a result, the test time
decreasesdrastically: instead of a dwell time equal to at least
one rotation of the mode stirrer in a conventionalreverberation
chamber,the dwell time in the intrinsic reverberationchamberis
equalto the dwell time of the EUT.
An intrinsic reverberationchamberhas beendrawn in Figure
6, side view, and Figure 7, top view.
The intrinsic reverberation room has been used by our
laboratory over a long period for acoustic investigations.The
dimensionsof the room are approximately 5.5x6x4.8m, with a
volume V=l64m” and a surface S= 180m’. The chamber is
constructed out of concrete. The walls will be lined with
copper foil in order to upgrade the chamber from acoustic to
electromagnetictest chamber. Test results were not available
before the publication date of this paper.
QUALITY FACTOR, ANALOGY ACOUSTICS
The most common tigure of merit when discussing
reverberation chambers is the Q of the chamber, which is
defined as the stored energy in the chamber divided by the
power that must be injected into the chamber,multiplied by the
angularfrequency,w=2rc$of operation.It is desirableto make a
ch,amberwith a Q as large as possible in order to give a high
field strength. The theoreticalexpressionfor the first-order Q’
of a rectanuularchamberis:
Q’=
3v
26
diffusers
floor
This was derived by writing the fields inside the chamberas a
seriesof cavity modesand then averagingover the ensembleof
all such modes after assumingthat an equal amount of energy
was in each mode. The modal method cannot be applied to
shapesother than those with separablegeometries[Dunn]. The
non-parallel walls of the intrinsic reverberation chamber
increasethe modal densitybut restrict the useof modal theory .
Dunn and Hill [Dunn, Hill 961 showed that for arbitrarily
shapedcavities an ensembleof plane waves can be used for
calculatingthe Q, resultingin the sameequation.The suggestion
on the use of non-parallelwalls for reverberationchambersfor
electromagnetictests is therefore also implicit in [Dunn, Hill
19961.
wall
L
.:
,P,S
volume[ m3]
S surface[m’]
Z$ skindepth[ml
ur relativepermeability
of wall
where:
Figure 6: Side view intrinsic reverberation chamber
V
In analogywith acoustics,wherethe reverberationtime TGcMB
is the figure of merit for a reverberationchamber.using Sabine’s
Equation:
acoustics(4)
where
TeorD
a
timewhereintheacousticnoise pressuredecreases
60dB. after stopping the noise source
absorptioncoefficient of the walls
we can also use the reverberationtime of an electromagnetic
reverberationchamber:
z=-= Q
fl
Figure 7: Top vieltt intrinsic reverberation chamber
0
35’
26,#,0 s
as shownby [Richiudsot~, Hill 961
cle~n~oll7~lglletics(5)
This relatioiship implies that one can determine the power
density in a reverberation chamber simply by measuringthe
reverberationtime and then measuringthe net power injected
into the chamber.It is not necessaryto use a calibratedreference
antennaor field probes.
CONCLUSION
A test technique for high power electromagnetic
immunity/susceptibilitytesting has been describedwhich makes
it possible to decreasetest time drastically: insteadof a dwell
time equal to at least one rotation of the mode stirrer in a
conventionalreverberationchamber,the dwell time in the novel
chamberis equal to the dwell time of the EUT.
REFERENCES/BIBLIOGRAPHY
In this short bibliographypublicationson reverberationrooms as
used in acoustics and reverberationchambers as used in
electromagnetics
areincluded.
Briiel & Kjar Technical Review no. 4 1978, ‘Reverberationat
Low Frequencies’H.
, Larsen
/SO/R 354-1963: ‘Measurementof absorptioncoefficients in a
reverberationroom’, dec. 1963.
IS0 354-1985: ‘Acoustics - Measurementof sound absorption
in a reverberationroom’, 1985.
IS0 3741-1988:‘Acoustics-Determinationof sound power
levels of noise sources-Precisionmethods for broad-band
sourcesin reverberationchambers’
MIL-STD-1344A, Method 1008, ‘Shielding Effectiveness of
Multicontact Connectors, Appendix A: Design of a ModeStirred Test Chamber’,10 Sept.1980
MIL-STD-1377 (Navy), ‘Effectivenessof Cables, Connectors,
and Weapons Enclosure Shielding and Filters in Precluding
Hazards of Electromagnetic Radiation in Ordnance;
Measurementof, Dep. of Defence,20 August 1971.
NBS Technical Note 1092, ‘Design, Evaluation and Use of a
Reverberation Chamber for Performing Electromagnetic
Susceptibility/Vulnerability Measurements’, M.L. Crawford,
G.H. Koepke,April 1986.
‘Modeling and
K.
Chamberlain,
B. Archambeault,
Measurementsof an Alternative Construction technique to
Reduce Shielded Chamber Resonance Effects’, IEEESymposiumon EMC, 1994
P.V. Briiel, Sound Insulation and Room Acoustics,
Chapman&Hall,19.5I
P. Corona, G. Latmiral, E. Paolini, ‘Performanceand Analysis
of a ReverberatingEnclosure with Variable Geometry’, IEEE
Transactionson ElectromagneticCompatibility, Febr. 1980,pp.
2-5
r\l.L. Crawford, G,H. Koepke, ‘OperationalConsiderationsof a
ReverberationChamber for EMC immunity MeasurementsSomeExperimentalResults’,IEEE Symposiumon EMC. 1984
J.M. Dunn, ‘Local, High-FrequencyAnalysis of the Fields in a
Mode-StirredChamber’,IEEE Transactionson Electromagnetic
Compatibility,Febr. 1990,pp. 53-58
M. Ha@eld (ed.), ‘Reverberation Chamber and Anechoic
Chamber Operators Group Meeting-Workshop’, 4 December
1995,NSWCDD, Dahlgren
M.O. Hageld, G.J. Freyer, M. B. Slocum, ‘Reverberation
Characteristicsof a Large Welded Steel Shielded Enclosure’,
IEEE Symposiumon EMC, 1997,pp. 38-43
D.A. Hill, ‘Electronic Mode Stirring for reverberating
Chambers’, IEEE Transactions on Electromagnetic
Compatibility,Nov. 1994,pp. 294-299
D.A. Hill, ‘A Reflection Coefficient Derivation for the Q of a
Transactions on
Reverberation Chamber’, IEEE
ElectromagneticCompatibility, Nov. 1996,pp. 59 l-592
V.O. Knudsen, Resonancein Small Rooms, J.A.S.A, vol. 4,
1932
J.M. L.&bury, G. H. Koepke, D.G. Camel& ‘Improvementsin the
CW Evaluationof Mode-Stirred Chambers’,IEEE Symposium
on EMC, 1997,pp. 33-37
F. B.J. Leferink, B. R.M. Pu$aert,
‘Accurate Shielding
EffectivenessMeasurementsUsing a ReverberationRoom, Int.
Symposiumon EMC, Zurich, 1993.pp 635-640.
T.A. Loughry, ‘FrequencyStirring, An Alternative Approach to
MechanicalMode-Stirring for the Conduct of Electromagnetic
SusceptibilityTesting’, USAF Phillips Laboratory, PL-TR-9l1036,November1991.
A.C.
Marvin,
J.A.S. Angus,
J. F.
Dawson,
J.
Clegg,
‘Enhancementsto Stirred Mode Chambers by the Use of
Pseudo-RandomPhasereflection Gratings’, Int. Symposiumon
EMC, Rome,Italy, 1994,pp. 218-221.
J. Clegg, A.C. Marvin, J.A.S. Angus, J. F. Dawson, ‘Optimal
PhaseReflection gratings and the Effect on Fields in a Mode
Stirred Chamber’,Int. Symposiumon EMC. Rome, Italy, 1996,
pp. 867-872.
A.K. Mittra, T.F. Trost, ‘Statistical Simulations and
MeasurementsInside a Microwave ReverberationChamber’,
IEEE Symposiumon EMC. 1997,pp. 45-53.
M. Petirsch, A. Schwab, ‘Optimizing Shielded Chambers
Utilizing Acoustic Analogies’, IEEE Symposium on EMC,
1997,pp. 154-158.
J.P. Quine, A.J. Pesta, ‘Stepped-FrequencyMethodology for
Obtaining faster Data Rates in Reverberation Chambers
Operated without a Mechanical Mode Stirrer’, International
Symposiumand Exhibition on EMC, Zurich 1995.
W.C. Sabine, ‘Collected Papers on Acoustics’. Harvard
University Press,1922(Also 196-l,Do\*erPublications)
R.E. Richardson. ‘Mode-Stirred Chamber Calibration Factor,
relaxation Time, and Scaling Laws’. IEEE Transactions on
ElectromagneticCompatibility, Dec. 1985,pp. 573-550
D.I. \v~r, D.C. Chang, ‘The Effect of an Electrically Large
Stirrer in a Mode-Stirred Chamber. IEEE Transactions on
ElectromagneticCompatibility. hluy 1989.pp. 164 I69