LEVEL 5 ELECTROMAGNETIC THEORY

AND RFID APPLICATIONS

Part 6: Electromagnetic compatiblilty measurements
Peter H. Cole
July 12, 2010

Contents
1 INTRODUCTION
1.1 Organisation of Part Six . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2 What EMC Means . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2 LINES PLANES AND ANTENNAS
2.1 Introduction . . . . . . . . . . . . . . . . . .
2.2 Ground Plane Eects . . . . . . . . . . . . . .
2.2.1 Charge and current images . . . . . . .
2.2.2 Monopole above ground plane . . . . .
2.2.3 Line parallel to ground plane . . . . .
2.2.4 Loop above ground plane . . . . . . . .
2.3 Coaxial lines . . . . . . . . . . . . . . . . . . .
2.3.1 Current flows . . . . . . . . . . . . . .
2.3.2 Generation of outside current . . . . .
2.3.3 Elimination of outside current . . . . .
2.3.4 Capacitive induction of current . . . .
2.4 Baluns . . . . . . . . . . . . . . . . . . . . . .
2.4.1 Transformer baluns . . . . . . . . . . .
2.4.2 Structural baluns . . . . . . . . . . . .
2.4.3 Equivalent circuit . . . . . . . . . . . .
2.5 Eect of Unscreened Components . . . . . . .
2.6 Practical EMC Measurement Antennas . . . .
2.6.1 Introduction . . . . . . . . . . . . . . .
2.6.2 Half wave dipole . . . . . . . . . . . .
2.6.3 Popular EMC antennas . . . . . . . . .
2.6.4 Antenna characteristics . . . . . . . . .
2.6.5 Calibration of antennas . . . . . . . . .
2.6.6 Antenna input and output descriptors
2.6.7 Conversion between terms . . . . . . .
3 EMC MEASUREMENT INSTRUMENTS
3.1 Introduction . . . . . . . . . . . . . . . . . .
3.2 Outdoor Test Site . . . . . . . . . . . . . . .
3.3 Indoor Test Site . . . . . . . . . . . . . . . .
3.4 Artificial mains network . . . . . . . . . . .
3.5 The Absorbing clamp . . . . . . . . . . . . .
3.6 Measuring Receiver . . . . . . . . . . . . . .
i

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.

1
1
1

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

3
3
3
3
4
5
6
6
6
7
8
8
10
10
11
15
15
17
17
17
18
18
20
20
20

.
.
.
.
.
.

21
21
21
23
23
25
26

ii

CONTENTS
3.7

The CISPR Quasi-peak Detector Receiver . . . . . . .

3.7.1 Evolution . . . . . . . . . . . . . . . . . . . . .
3.7.2 Frequency bands . . . . . . . . . . . . . . . . .
3.7.3 Filter section . . . . . . . . . . . . . . . . . . .
3.7.4 Detector section . . . . . . . . . . . . . . . . . .
3.7.5 Indicator section . . . . . . . . . . . . . . . . .
3.7.6 Summary of parameters . . . . . . . . . . . . .
3.7.7 Output variation with impulse rate . . . . . . .
3.7.8 Interior waveforms . . . . . . . . . . . . . . . .
3.7.9 Response to isolated impulse . . . . . . . . . . .
3.7.10 Response to continuous wave signals . . . . . .
3.7.11 Response to modulated continuous wave signals
3.7.12 Factors regulating output . . . . . . . . . . . .

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

27
27
27
28
28
30
30
32
33
33
36
36
36

List of Figures
1.1

Sources and suerers of electromagnetic interference. . . . . . . . . . . . .

2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
2.10
2.11
2.12
2.13
2.14
2.15
2.16
2.17
2.18

Charge and current images in a conducting plane. . . . . . . . . . .

Monopole antenna and its image. . . . . . . . . . . . . . . . . . . .
Line antenna parallel to groudn plane. . . . . . . . . . . . . . . . .
Loop antenna above ground plane. . . . . . . . . . . . . . . . . . .
Currents on a co-axial line. . . . . . . . . . . . . . . . . . . . . . . .
Inadvertant generation of currrent on the outside of a co-axial line.
Elimination of currrent on the outside of a co-axial line. . . . . . . .
Three-terminal capacitor. . . . . . . . . . . . . . . . . . . . . . . . .
Unbalanced feed to a balanced dipole. . . . . . . . . . . . . . . . . .
Unbalanced feed to a balanced loop antenna. . . . . . . . . . . . . .
Balanced feed to a balanced dipole antenna. . . . . . . . . . . . . .
Balanced feed to a balanced loop antenna. . . . . . . . . . . . . . .
A structural balun for a balanced dipole. . . . . . . . . . . . . . . .
A structural balaun for a balanced loop. . . . . . . . . . . . . . . .
Equivalent circuit for feed and loop. . . . . . . . . . . . . . . . . . .
Analysis of undscreneened components. . . . . . . . . . . . . . . . .
Antenna pattern for half wave dipole. . . . . . . . . . . . . . . . . .
A range of EMC measurement antennas. . . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

4
5
5
6
7
8
8
9
9
10
11
12
13
14
14
16
18
19

3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
3.10
3.11
3.12
3.13
3.14

Plan view of CISPR outdoor test site. . . . . . . . . . . . . . . . . . .

Elevation of a CISPR outdoor test site. . . . . . . . . . . . . . . . . .
CISPR test site for conducted measurements. . . . . . . . . . . . . .
Circuit for artificial mains V-network. . . . . . . . . . . . . . . . . . .
Construction and usage of the absorbing clamp. . . . . . . . . . . . .
Simplified view of an emc receiver. . . . . . . . . . . . . . . . . . . .
Limits of overall selectivity (Band C). . . . . . . . . . . . . . . . . . .
Filter response for a quasi-peak detector receiver. . . . . . . . . . . .
Detector circuit for a quasi-peak detector receiver. . . . . . . . . . . .
Circuit for simulating a moving coil meter. . . . . . . . . . . . . . . .
Summary of parameters for the quasi peak detector receiver. . . . . .
Response of quasi-peak detector receiver for various impulse rates. . .
Waveforms in quasi-peak detector receiver for standard impulse rates.
Indicator response to infrequent impulses. . . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.

22
23
24
24
25
26
28
29
29
30
31
33
34
35

iii

iv

LIST OF FIGURES

Chapter 1
INTRODUCTION
1.1

Organisation of Part Six

Part six contains only three chapters devoted to the study elelectromagnetic compatiblity
and to electromagnetic compatibility measurements as they aect RFID systems.
As an important aspect of RFID systems is their need to satisfy electromagnetic
compatibility regulations in the jurisdiction in which they operate, and such regulations
are written in terms of measurements using test equipment specialised for Electromagnetic
Compatibility regulations, this Part of the course includes a chapter on the construction
and behaviour such test equipment.

1.2

What EMC Means

The subject of electromagnetic compatibility is the study of the ways in which dierent
users of the electromagnetic spectrum can happily co-exist without producing unacceptable interference to the performance of one anothers systems. This involves
Studying how such interference can come about.
Studying how the eects can be measured.
Studying how regulations may be devised to control activities which might produce
interference.
Studying the steps that may be taken to minimise such interference.
An illustration of many aspects of electromagnetic interference in the everyday environment is shown as Figure 1.1. In the diagram we should recognise that interference
may be generated
by man made devices, or
by nature
and that it may be
radiated to, or
1

CHAPTER 1. INTRODUCTION

Figure 1.1: Sources and suerers of electromagnetic interference.

cable borne to
things that may suer from interference.

Chapter 2
LINES PLANES AND ANTENNAS
2.1

Introduction

Now we have mastered the elementary principles of radiation, we turn to a consideration of practical issues in applying radiation concepts in the context of electromagnetic
compatibility. Issues we will have to consider are
Some of the peculiarities of transmission lines, particularly mismatched or unmatched lines.
The eect of ground planes on the fields created by antennas placed in close proximity to them.
The properties of practical antennas used for electromagnetic compatibility measurements, and the engineering terms by means of which their properties are described.
This last need arises from the fact that the small electric and magnetic dipole antennas
considered in previous parts of the course have the property that they are both inecient
and of narrow bandwidth.

2.2
2.2.1

Ground Plane Eects

Charge and current images

We begin with the study of ground planes. When a system of charges and currents is
placed adjacent to a ground plane, the field which would be generated (in the absence of
the ground plane) does not satisfy the requirements of zero tangential electric field and
zero normal magnetic flux density which are characteristic of conducting planes. What
then happens is that a system of charges and currents is induced in the ground plane so
that the total field which arises from the original charges and currents and the ground
plane charges and currents does satisfy the boundary conditions.
The method of images, which provides a simple way to calculate the eect of the
ground plane charges and currents, is illustrated in Figure 2.1. There a charge q above a
ground plane is said to produce an image charge q an equal distance below it, while a
current element idr above the ground plane provides an image current of equal magnitude
(but not the same direction) an equal distance below it. The direction of the image current
3

CHAPTER 2. LINES PLANES AND ANTENNAS

Figure 2.1: Charge and current images in a conducting plane.

element is such that the components of the original and image element perpendicular to
the ground plane are similarly directed, while the components of the original and image
current elements parallel to the plane are oppositely directed.
It is clear that the total field which would be created by the original charge and current
and their images, if both were present and the conducting plane absent, has a symmetry
which does satisfy the boundary conditions viz . zero tangential electric field and normal
magnetic flux density which are required at the plane. What is also true is that, above the
ground plane the field is the same as is caused by the original charge and current together
with whatever charges and currents flow on the conducting plane. Of course below the
ground plane, the field created by the original charge and current, and the ground plane
charge and current is zero, as the ground plane eectively shields the region below it from
the eects of the original sources.
This method of images provides a simple way of calculating fields and sources in the
presence of ground planes: we just introduce the images and forget about the ground
plane.

2.2.2

Monopole above ground plane

Applying this notion we see that a monopole above a ground plane produces an image
as shown in Figure 2.2. If the monopole is fed at its base by a voltage generator V , the
image will be fed by an image generator also of magnitude V , with polarity as shown. Of
course the image generator does not exist, and the implied power which it would provide
does not exist, because below the ground plane the power which would be provided by
both the original generator and the image generator does not exist.
One thing we might note is that monopole above ground produces the same field as a

2.2. GROUND PLANE EFFECTS

Figure 2.2: Monopole antenna and its image.

dipole in free space. If the dipole is made of a resonant length /2, the monopole would
be resonant at a length of /4.

2.2.3

Line parallel to ground plane

Figure 2.3: Line antenna parallel to groudn plane.

A straight wire parallel to a ground plane and its image are shown in Figure 2.3. Again
a real voltage generator above the ground plane produces an image generator below.
The eect of the image in this case stands in significant contrast to the previous case
of a wire perpendicular to the ground plane. In the former case the wire and image
formed a structure, (namely a wire antenna in the form of an electric dipole), similar to
the original structure. In that case the presence of the ground plane will enhance the
radiation. In the present case the direction of the current in the image is opposite to
that of the original current, and if the original wire is close to the ground plane, the
combination is best viewed as a loop antenna, a structure very dierent from that of the
original wire antenna, and of significantly reduced radiation relative to that of the original
wire antenna.

CHAPTER 2. LINES PLANES AND ANTENNAS

2.2.4

Loop above ground plane

Figure 2.4: Loop antenna above ground plane.

This eect is further illustrated in Figure 2.4, which shows the original and image for
a wire loop which lies close to and parallel with the ground plane. When we take note
of the relative senses of current elements in the original loop and its image, we see that
if the separation is much less than a wavelength, a substantial tendency for the radiation
to cancel occurs.

2.3

Coaxial lines

It is now necessary to take note of some basic facts about currents which can flow on
transmission lines and in particular of the important eects of currents which may flow
on the outside of coaxial lines.

2.3.1

Current flows

At high frequencies, currents always flow on the outside of metal surfaces, and do not
significantly penetrate metal, and certainly do not cross to the other side. In consequence,
assuming the wall thickness of a coaxial cable is adequate to perform its intended screening
function, we can identify as shown in Figure 2.5 for a coaxial line three independent
surfaces on which currents can flow.
The outside surface of the outer conductor.
The inside surface of the outer conductor.

2.3. COAXIAL LINES

The outside surface of the inner conductor.

Figure 2.5: Currents on a co-axial line.

The Figure also illustrates an important fact about the relationship between these
currents which is that, inside the line, the properties of TEM modes which fill the line
ensure that the forward and return currents are equal. Thus inside the line, the total
current (shown as ii in the figure) returning on the inside of the outer conductor is the
same in magnitude (but in the opposite direction) as the total current which flows on the
inner conductor.
The current which flows on the outside of the outer conductor and shown as ie in
Figure 2.5 is, however, entirely independent of the current which flows inside. Such outer
current is termed `common mode, as it is, in view of the fact that the two currents
inside sum to zero, equal to the sum of all three currents. When the common mode
current is present, an interaction, often through radiation but often also through a direct
connection, between that current and the outside world, i.e. the part of the world outside
of the exterior of the coaxial line and the interior of any screened structures to which it is
connected, takes place. In an EMC context, such interaction is almost always undesired.

2.3.2

Generation of outside current

A method by means of which such outside currents can be, in this case inadvertently,
generated is illustrated in Figure 2.6. Here a voltage generator Vs of internal impedance
Zs in a screened box drives current i into the centre conductor of a coaxial line which
enters the screening box through a hole, but makes no contact with the screening box as
it does so. The relation between currents inside the cable causes a current i to flow on
the inside of the outer conductor in response to the current i on the centre conductor.
The current i, when it encounters the end of the cable inside the box, has nowhere to
flow other than onto the outside of the outer conductor where it becomes a current +i
thereon.

CHAPTER 2. LINES PLANES AND ANTENNAS

Figure 2.6: Inadvertant generation of currrent on the outside of a co-axial line.

2.3.3

Elimination of outside current

Figure 2.7: Elimination of currrent on the outside of a co-axial line.

Assuming as is invariably the case the outside current is undesirable, the remedy is
shown in Figure 2.7. The outer conductor of the coaxial line makes contact with the
screening box as it enters therein, most desirably at points all around the circumference
of the cable, so there is no possibility that the currents already launched on the outside
of the cable can emerge from the screening box. These currents remain on the inside of
the screening box, and return to the voltage generator via the path shown. The part of
the coaxial cable on the outside of the box carries only the equal and opposite interior
currents as is generally intended.

2.3.4

Capacitive induction of current

A second mechanism by means of which currents may be induced onto the outside of
coaxial lines is through their proximity to sources of unbalanced electric potential.
To understand this eect we remind ourselves that in the presence of ground, a full
expression of capacitance say between two nodes labelled as A and B in Figure 2.8 includes
as well as the direct capacitance CAB between nodes A and B, two additional capacitances
CAE and CBE between these nodes and ground. The structure illustrated is known as
a three-terminal capacitor, and we should always be prepared to look, in addition to

2.3. COAXIAL LINES

Figure 2.8: Three-terminal capacitor.

any impedances overtly illustrated between any nodes of a circuit, for stray capacitance
between such nodes, and for capacitance between those nodes and the ground.

Figure 2.9: Unbalanced feed to a balanced dipole.

The eect of such capacitances, and their ability to induce currents on the outside of
a coaxial line, is illustrated in Figure 2.9 in which a simple minded attempt to establish
a connection between a coaxial feed line and a centre fed dipole is made. Both arms
of the dipole have the same capacitance to a nearby parts of the coaxial feed line, but
for only one arm is there a significant potential dierence. Thus from one arm there
is a capacitive current induced on the outside of the cable, but from the other arm no
compensating current with opposite sign to cancel it exists. We are left with a net current

10

CHAPTER 2. LINES PLANES AND ANTENNAS

on the outside of the cable with generally undesirable eects.

Figure 2.10: Unbalanced feed to a balanced loop antenna.

The same problems are evident in Figure 2.10, in which a similarly unsuccessful (in
the sense that currents on the outside of the feed line result) attempt is made to excite a
balanced loop antenna from a coaxial line.

2.4
2.4.1

Baluns
Transformer baluns

An appropriate way to deal with the problem just described lies in the use of balanced to
unbalanced transformers generally known as baluns, as is shown in Figure 2.11.
The balun is physically small, so that its own capacitances are not significant, and
may have an electrostatic screen between primary and secondary winding so that coupling
between those windings is entirely through magnetic field. One winding (the unbalanced
side) is connected to the end of the coaxial feed line. The other winding (the balanced side)
has a centre tap, and the outer ends of that winding develop equal-magnitude oppositephase voltages with respect to that centre tap. The centre tap is connected to the outside
of the coaxial cable, with the result that the capacitively induced currents thereon from
the two arms of the dipole are equal and opposite, and thus sum to zero.
Figure 2.12 shows a balun being successfully used to excite a loop antenna without a
net induction of current on to the outside of the cable.

2.4. BALUNS

11

Figure 2.11: Balanced feed to a balanced dipole antenna.

2.4.2

Structural baluns

Not all baluns involve the use of a transformer. Figure 2.13 shows an example of what
might be called a structural balun in which the structure of the environment outside of
the coaxial cable and near to it is modified so that, despite the unbalanced nature of the
geometry inside the cable, the exterior region has a balanced symmetry with respect to
ground (which may be near or far) and with respect to the two arms of the balanced
structure being fed, in this case a centre-fed dipole.
In this structure, we can recognise two sets of transmission lines. One is formed by
the conducting surfaces inside the coaxial line. This is an unbalanced line, but is screened
from the outside world by the outer conductor of that line. The other is a twin-wire line
and is formed by the outside of the outer conductor of the coaxial line and the outside of
the outer conductor of another (and possibly solid) dummy line of the same dimensions.
These two lines are joined (i.e. shorted) at a point /4 from the feed point of the
dipole. The short circuit at that point produces an open circuit (for this line) at the point
of connection to the arms of the dipole, so that no dierential mode current enters at that
point the transmission line formed by the outers of the two conductors.
So far as common mode current flowing on that twin line is concerned, examination of
the structure is such as to assure us that the geometric symmetry between that line and
the dipole arms is such that no common mode current will be induced. Put another way,
the current which the right hand dipole arm will induce (through capacitance) on to the
left hand outer conductor is equal in magnitude but opposite in sign to the current which
will be induced on the right hand outer conductor by the left hand dipole arm. When
these equal and opposite currents combine at the junction point of the twin wire line they
cancel, and there is no surviving current to proceed further along the side of the coaxial
line away from the dipole.

12

CHAPTER 2. LINES PLANES AND ANTENNAS

Figure 2.12: Balanced feed to a balanced loop antenna.

2.4. BALUNS

13

Figure 2.13: A structural balun for a balanced dipole.

Attractive as this structure is, it is not a solution for all frequencies. It works for
frequencies for which the length shown as /4 is or is close to a quarter wave length. At
other frequencies the performance begins to deteriorate.
Another structural balun, this time without that frequency limitation, is shown feeding
a loop antenna in Figure 2.14. Here the loop has a feed point at the gap shown, and a
conductor placed directly opposite provides physical support, or a means of holding the
loop, and may eventually lead to ground. Clearly from a point of view which considers
only the outside of the conductors (and ignores the small asymmetry at the gap), the
structure is symmetrical and balanced with respect to ground: the two sides of the loop
will induce cancelling currents on the outside of the feed cable.
What is notable about this structure is that half of the loop and the feed cable have
been hollowed out so that the coaxial line may reach the feed point at the gap in the
loop without disturbing the balanced structure just discussed. We must remember that
the inside of the feed cable and the outside of the feed cable are two separate worlds in
which independent field solutions may exist: balanced outside the cable and unbalanced
inside the cable. The only connection between them is at the feed point of the loop, but
the geometry here is considered to be suciently small for it to produce but a minor
perturbation to the balanced nature of the loop as a whole.

14

CHAPTER 2. LINES PLANES AND ANTENNAS

Figure 2.14: A structural balaun for a balanced loop.

Figure 2.15: Equivalent circuit for feed and loop.

2.5. EFFECT OF UNSCREENED COMPONENTS

2.4.3

15

Equivalent circuit

An equivalent circuit for the loop and feed line for the structure just discussed is shown
in Figure 2.15. The left hand section consisting of a transmission line of characteristic
impedance Z0 and length L1 and propagation constant (inside the cable) of 1 represents
the feed line from its point of origin up to the loop gap. The lumped circuit inductance
LE represents the eective inductance of the loop, in which current flows on the outside
surface.
The length L2 of transmission line of characteristic impedance Z0 and propagation
constant 1 , terminated in a short circuit on the right hand side of the circuit, represents
the right hand side of the loop, assuming it is also hollow for some distance beyond the
feed point. Frequently, and indeed as shown in Figure 2.15, that half of the loop is solid,
and L2 in Figure 2.15 should be set to zero.

2.5

Eect of Unscreened Components

Yet another example of the way in which currents can be induced onto the outside of
coaxial lines is when they are used to feed physically large unscreened structures with
capacitance to ground. This case will be illustrated by a particular example illustrated in
Figure 2.16(a). Here a coaxial line which originates in a screened and grounded voltage
source feeds, without the benefit of a balun, a pair of nodes A and B between which there
is a load ZL . The two nodes might for example be the chassis of an unscreened instrument
(node A) upon which there is mounted some unscreened bulky component (node B). The
chassis has a capacitance C2 to ground and the bulky component has at its node B side
a capacitance C1 to ground.
In analysis of this arrangement, we will assume the dimensions of the system are
small enough in relation to the wave length at the operating frequency for lumped circuit
approximations to apply, and the capacitances C2 and C1 terminate at points suciently
close to one another on the ground for that region to be considered as a single node.
We will also assume that the screened source and length of coaxial line is such as to
produce, looking into the end of the coaxial line which feeds nodes A and B, a source
voltage VS behind an internal impedance ZS . That arrangement feeds the load impedance
ZL so that the source system which feeds the system of stray capacitances C1 and C2 and
cable external inductance (discussed below) is as shown in Figure 2.16(b).
If we simplify using Thevenins theorem the source system just discussed, and add to
the circuit the two capacitors C1 and C2 , and the inductance which is defined by the path
starting at node E, moving leftward along the ground to the region of the source system
box, following the outside of the coaxial cable to node A, then through the capacitance
C2 to the starting point E we have an equivalent circuit as shown in Figure 2.16(c).
In that equivalent circuit we can write for the eective inductance the formula
}

0 D
8D
L=
loge
2
4
d

(2.1)

in which we have made use of a well known result for the inductance of a single-turn
circular coil, but have adjusted that result here to account for the fact that the field
created by the semi-circular coil in the presence of its image in the ground plane is the

16

CHAPTER 2. LINES PLANES AND ANTENNAS

Figure 2.16: Analysis of undscreneened components.

2.6. PRACTICAL EMC MEASUREMENT ANTENNAS

17

same as that which would be produced if the full coil were present, but the flux linked by
the semi-circular path is half that of that full circular coil.
Because the circuit has become complex, we are now going to make some simplifications, namely that the impedance of the source system feeding nodes A and B is low
enough for us to set it by way of approximation to zero, so we have a source of magnitude
which we will now for convenience called V and no internal impedance feeding nodes A
and B, as shown in Figure 2.16(d). We must admit that this is a significant and maybe
inaccurate assumption, in that it removes from the circuit any damping to any resonant
circuits that may be present. It is made mostly for the convenience in analysis that it
aords. We should continue to bear in mind its possible inaccuracy. Looking now at the
rest of the circuit from the point of view of the inductor, we may again use Thevenins
theorem to produce the simpler circuit shown in Figure 2.16(e).
From this simplified equivalent circuit we can easily calculate the current which flows
on the outside of the coaxial line as the current flowing in this circuit through L.
Although we will not do this explicitly, we may draw some conclusions about the
process. The first is that the current depends for its existence on a non-zero value for C1 .
The existence of the capacitance C2 alone is not sucient for that current to flow. The
capacitance C2 does, however, have a role to play in establishing the magnitude.
The second thing we notice is that the circuit contains the possibility of resonance. If
the inductance L approximately tunes the parallel combination of C1 and C2 , the currents
on the outside of the cable could become large, at least under the assumption we have made
earlier of negligible source impedance at the output of the coaxial line. Such resonances
are entirely possible when an unscreened bulky equipment is fed over significant coaxial
line lengths which are raised above ground so that a significant inductance producing flux
collecting area is created.

2.6
2.6.1

Practical EMC Measurement Antennas

Introduction

The previous chapter considers the properties of elementary and small electromagnetic
antennas. Now we consider just some of the properties of more developed antennas
suitable for measurement purposes.
Those antennas considered previously have a property in common with all small antennas, namely that the radiation resistance is low or very low, and is accompanied by large
reactance, so antennas for practical use require tuning to remove the reactance to allow
ecient power transfer. The tuning elements themselves have losses and may attribute to
loss of eciency. Even if the loss of eciency is not severe, the resulting resonant circuit
is of low bandwidth. We now consider qualitatively the range of larger antennas free of
those particular limitations.

2.6.2

Half wave dipole

An ecient antenna of convenient source or driving point impedance is made by a centre

fed half wave dipole. This means the antenna is intended to operate at or near the
frequency for which its total length is half of a wave length.

18

CHAPTER 2. LINES PLANES AND ANTENNAS

Figure 2.17: Antenna pattern for half wave dipole.

This antenna has an approximately sinusoidal current distribution, beginning with zero
at one end, rising to a maximum at the centre, and falling to zero at the other end. The
antenna pattern diers only slightly from that of the short electrical or magnetic dipole,
with the result that the gain factor (assuming no losses) is 1.64. A wire-frame illustration
of the antenna pattern is given in Figure 2.17. The input (and source) impedance of the
/2 dipole can be shown to be 72 ohms, with no accompanying reactance at the centre
frequency and only small reactances at nearby frequencies. It is a very convenient antenna
for operation over a limited frequency range. Because, however, EMC investigations
normally cover very wide frequency ranges, a range of broader band antennas shown in
Figure 2.18 is frequently used.

2.6.3

Popular EMC antennas

A collage of popular EMC antennas made by a single manufacturer is shown in Figure 2.18.
The photograph shows, inter alia, (a) the half wave dipole just discussed, (b) a biconical
antenna, (c) a short vertical monopole with preamplifier in the box beneath it, (d) a
screened loop antenna with preamplifier in the box beneath it, (e) a log periodic dipole,
(f) a coaxial two-arm spiral antenna, and (g) several small loop antennas for near field
coupling. We will not record in detail here the properties of these antennas but some
discussion will be provided in other sections.

2.6.4

Antenna characteristics

All of these antennas are normally used with calibration factors supplied by the manufacturers. The calibration factors express the output voltage to a specified load (generally
50 ohms) for a given incident plane wave, the strength of which is expressed in terms of
either its electric field or its magnetic field.

2.6. PRACTICAL EMC MEASUREMENT ANTENNAS

Figure 2.18: A range of EMC measurement antennas.

19

20

2.6.5

CHAPTER 2. LINES PLANES AND ANTENNAS

Calibration of antennas

The calibration is generally based on the reciprocity method, i.e. the propagation loss
between identical antennas is measured, minimising or taking into account the eects
of ground reflections and using the theoretical relationships between propagation loss,
antenna gain and propagation distance as given in earlier parts of the course.

2.6.6

Antenna input and output descriptors

In the expression of antenna input and output, logarithmic variables are used. Antenna
output is usually expressed as dB relative to one watt (dBW) or relative to one milliwatt
(dBmW), or relative to one microvolt, i.e. as dBV. For antennas which are basically
electric field sensors the incident field is expressed in dB relative to one volt per metre,
i.e. dBV/m or dB relative to one microvolt per metre, i.e. dBV/m. Incident field for
basically magnetic field sensitive antennas is expressed in dB relative to one amp per
metre, i.e. dBA/m or dB relative to one microamp per metre, ie dBA/m.

2.6.7

Conversion between terms

Care must be taken in converting between the forms of expression of antenna output and
antenna incident field. Such conversions are unavoidable in practical work.
The general rule to remember is that dB are intended to provide an expression of a
power ratio, and if powers are being compared one takes 10 log to the base 10 of the power
ratio, while if voltages are being compared one takes 20 log to the base 10 of the voltage
ratio.
In translating antenna output voltage to antenna output power, in an assumed 50 ohm
system, we deduce from the equation Pout = |Vout |2 /50, and the fact that 10 log to the
base 10 of 1/50 is -17, that in a 50 ohm system
Power in dBW = voltage in dBV 17

(2.2)

.
In translating antenna incident fields from electric field terms to magnetic field terms,
and on the assumption that the fields have the ratio of characteristic impedance of free
space, we may write, in view of the fact that 20 log10 120 = 51.5
Field in dBA/m = Field in dBV/m 51.5.

(2.3)

In comparing the last two results it is interesting to note that in one case the constant
to be subtracted is 10 log to the base 10 of an impedance, while in the second the constant
to be subtracted is 20 log to the base 10 of an impedance. Strange but true.

Chapter 3
EMC MEASUREMENT
INSTRUMENTS
3.1

Introduction

Electromagnetic compatibility measurements require the use of the range of specialised

test instruments itemised below.
For interfering signal measurements, a measurement receiver or alternatively or
additionally a suitably configured spectrum analyser.
For radiated interference measurements, an outdoor test site or a calibrated anechoic
chamber, and a range of suitable calibrated antennas.
For conducted (i.e. cable borne) interference measurements, a suitably configured
indoor test site.
For some conducted interference measurements, an Artificial Mains Network, alternatively known as a Line Impedance Stabilisation Network (LISN).
For other conducted interference measurements, an absorbing clamp.
The properties of suitable antennas have been outlined, somewhat briefly, in the previous chapter. The properties of the remaining instruments will be outlined in the sections
to follow.

3.2

Outdoor Test Site

Essential characteristics of the CISPR outdoor test site for radiated measurements are
illustrated in Figure 3.1 and Figure 3.2.
Things which might be noted are:
The dimension D is the measurement distance and varies in the regulations with
frequency and regulatory authority.
The size of the CISPR ellipse which must be kept clear of extraneous objects is
determined by the dimension D in such a way that any reflections from such objects
involve a propagation path of at least 2D.
21

22

CHAPTER 3. EMC MEASUREMENT INSTRUMENTS

Top View
Major Diameter = 2D

Minor
Diameter
= 3D
__
D
2

__
D
2

Equipment Under
Test (EUT)

Measuring
Antenna

Coaxial
Cable

Spectrum Analyzer
or Field-strength
Meter

Minimum area(ellipse)
to be free of reflecting
objects
D=3m, 10m,
30m (Typically)

Figure 3.1: Plan view of CISPR outdoor test site.

3.3. INDOOR TEST SITE

23

Figure 3.2: Elevation of a CISPR outdoor test site.

The site has a conducting ground plane made from a metal mesh of pitch not greater
than one tenth of a wavelength at the highest operating frequency.
The device under test is rotatable through all angles.
The pickup antenna is adjustable in height and in polarisation.

3.3

Indoor Test Site

Essential characteristics of the CISPR indoor test site for conducted measurements are
illustrated in Figure 3.3.
Things which might be noted are:
There is an implied ground plane at floor level.
There is an additional vertical ground plane behind the instrumentation.
There are particular requirements defined in the text of the regulations about the
positioning of cables.

3.4

Artificial mains network

The circuit diagram of the CISPR artificial mains V-network for conducted interference
measurements is illustrated in Figure 3.4. The basic properties of the instrument are:
It prevents mains borne noise from the test environment from reaching the device
under test or the test receiver.
The radio frequency impedance presented to the mains input of the device under
test is standardised.

24

CHAPTER 3. EMC MEASUREMENT INSTRUMENTS

Figure 3.3: CISPR test site for conducted measurements.

Figure 3.4: Circuit for artificial mains V-network.

3.5. THE ABSORBING CLAMP

25

Radio frequency interference emerging from the device under test via its mains cable
is coupled to the measuring receiver for measurement of its amplitude.
Independent measurement of the interference on the active and neutral power lines
is catered for.

3.5

The Absorbing clamp

The Absorbing Clamp is used as an indirect indicator of currents flowing along the mains
lead of a device (with clamp in place), for frequencies in the range 30 MHz to 300 MHz.
The construction and usage of the Absorbing Clamp is illustrated as Figure 3.5.

Figure 3.5: Construction and usage of the absorbing clamp.

CISPR experience shows that the disturbing energy from a device is mostly radiated
from the portion of the mains lead near the appliance. The disturbing capability of a
device is defined by CISPR to be the power it could supply to its mains lead. This power
is said to be nearly equal to that supplied by the appliance to a suitable absorbing device
placed around this lead at the position where the absorbed power is at a maximum.
Some interesting observations on using the clamp are:
The clamp is slowly adjusted along the mains lead at each test frequency so that
the maximum indicated output is found.
A current transformer is used to provide an input to the CISPR measurement receiver.
Lossy ferrite sleeves are used to attenuate the current flowing further along the
mains cable and to provide a better impedance match.

26

CHAPTER 3. EMC MEASUREMENT INSTRUMENTS

The clamp also prevents high frequency interference from elsewhere in the test site
from reaching the sample point.
The clamp measurements are calibrated by a substitution technique:
A signal generator with a matched 50 ohm output impedance is substituted for the
DUT
The output of the generator is adjusted to give the same reading from the clamp as
obtained with the DUT.
The signal generator output is then taken as the power which would be delivered to
mains but with the clamp absent.

Clearly some assumptions and approximations are required in justifying the above
statements.

3.6

Measuring Receiver

Figure 3.6: Simplified view of an emc receiver.

A simplified view of the contents of an emc measuring receiver is shown on Figure 3.6.
The basic ingredients are:
A tuneable bandpass filter with a precisely defined bandwidth and skirt steepness.
A detector circuit to convert some aspect (e.g. peak value, envelope amplitude,
amplitude, average power etc) of the radio frequency signal emerging from the filter
to a low frequency which may be indicated by a simple (e.g. moving coil) meter.
The simple indicating meter which is read by the user.
For various types of measurements, dierent bandwidths and skirt steepness of the
bandpass filter are required. For example, measurement of adjacent channel interference
in a multi channel transceiver (eg mobile radio) requires a very high steepness of the skirts
of the filter response. Contrastingly, measurements of the capacity of various signals to
interfere with broadcast radio requires that the bandpass filter mimic that found in simple
radio receivers, and be of only moderate skirt selectivity.
For various types of measurements the nature of the detector also varies. Four types
of detector which are in common use are:

3.7. THE CISPR QUASI-PEAK DETECTOR RECEIVER

27

The peak detector , which produces an output proportional to the peak value of
the waveform emerging from the filter, and ignores variation in the shape of the
envelope.
The average detector , which measures the long time average of the amplitude of
the envelope of the signal emerging from the filter.
The rms detector , which provides an indication of the long time average of the power
in the signal emerging from the filter.
The quasi-peak detector, the output of which has complex relation to the envelope
of the signal emerging from the detector and is explained in more detail in sections
below.
Of particular importance is the quasi-peak detector, which is used in a very large number of CISPR measurements, yet appears to be insuciently understood for its response to
arbitrary waveforms to be easily foreseen by emc practitioners. We will therefore consider
it properties in some detail in following sections.

3.7
3.7.1

The CISPR Quasi-peak Detector Receiver

Evolution

The origins of the quasi-peak detector receiver lie in the early days of broadcasting, and in
the observation that interference in the form of repeated very short (i.e. extremely broad
band) but infrequent impulses becomes, to human users of the broadcast system, less
objectionable as the repetition rated frequency of these impulses reduces. The quasi-peak
detector receiver was devised to produce an output signal which was believed to mimic
the perception of most humans of the degree to which impulses of dierent repetition
rates are objectionable to them. Since its initial development, the application of the
quasi-peak detector receiver has been extended to contexts and frequency bands other
than originally used for broadcasting. The extension has arisen more as a convenient
measurement convention than because of any proven capacity of the quasi-peak detector
receiver to model the objectionable behaviour of impulses in the broader contexts and
frequency ranges in which it is now applied. The specifications of the receiver employing
the quasi-peak detector extend to all three blocks of the basic diagram in Figure 3.6, and
in addition vary with intended operating frequency.

3.7.2

Frequency bands

For the purpose of definition, four frequency bands are recognised. These are: Band A
- 9 kHz to 150 kHz; Band B - 150 kHz to 30 MHz; Band C - 30 MHz to 300 MHz; and
Band D - 300 MHz to 1000 MHz. These parameters will be listed, together with other
parameters defining the quasi-peak detector for the dierent bands, further below.

28

3.7.3

CHAPTER 3. EMC MEASUREMENT INSTRUMENTS

Filter section

In each of these bands the receiver has a specified bandwidth and has a specified skirt
selectivity, both of which vary for each band. The specified limits for the skirt selectivity
for Band C are illustrated in Figure 3.7.

Figure 3.7: Limits of overall selectivity (Band C).

An analytic model of the filter response that is often employed to perform the calculations, described later, of the response to repeated impulses is illustrated as Figure 3.8.
This response is obtained from the cascade of two critically coupled double tuned circuit
responses arranged to produce the desired passband at 6 dB.

3.7.4

Detector section

The detector section of the quasi-peak detector receiver has a circuit diagram as shown in
Figure 3.9. In that circuit, the charging time of the reservoir capacitor C is limited by the
resistor R1 , while the discharging time is limited by the resistor R2 . Because the charging
of the capacitor takes place only during a portion of the positive half cycles of the a.c.
waveform emerging form the filter, the time constant of charging of the capacitor when an
a.c. signal is suddenly applied to the receiver is not the product R1 C; numerical analysis
shows that the time to charge to 63% of final value under those conditions is about four
times the product R1 C. The discharging of the reservoir capacitor, however, takes place
between impulses in a simple exponential way, and the discharging time constant is the
same as the product R2 C.

3.7. THE CISPR QUASI-PEAK DETECTOR RECEIVER

Figure 3.8: Filter response for a quasi-peak detector receiver.

Figure 3.9: Detector circuit for a quasi-peak detector receiver.

29

30

3.7.5

CHAPTER 3. EMC MEASUREMENT INSTRUMENTS

Indicator section

The indicator section of the quasi-peak detector receiver takes the form of a critically
damped second order response system of specified time constant, and may be realised by
a simple lightweight moving coil meter, or by an electronic circuit which has the same form
of response. The time constant of the meter varies somewhat between bands, and will be
specified in the summary of parameters further below. A simple RLC circuit simulating
(if such a meter is not used) this response is shown in Figure 3.10.

Figure 3.10: Circuit for simulating a moving coil meter.

3.7.6

Summary of parameters

Figure 3.11 contains for the four bands defined earlier, a summary of the parameters which
define the quasi-peak detector receiver. We remark again that the parameter shown as
charging time is not the product R1 C, but is about four times that value, a numerical
analysis being required to determine the exact relation, which varies from band to band.
The Figure also shows that the time constant of the second order response of the indicating
meter is characteristic of a lightweight mechanical system.
The Figure also shows some parameters which have not yet been discussed. Three
of these reflect the fact that the quasi-peak detector receiver was devised to serve as
an indicating instrument for a repeated sequence of impulses. In each band, there is a
standard impulse size, measured in volt seconds, and a standard repetition rate, measured
in pulses per second, for which the receiver is supposed to produce a particular standard
output. The standard impulse sizes quoted are internal to the impulse generator, and it
is assumed that the generator has an output impedance matching the input impedance of
the measuring instrument, so the impulses seen at the input terminals to the quasi-peak
detector receiver have an amplitude half that appearing in the Figure.
The standard output is whatever is produced by the same receiver when it has, as its
input, a 1 mV r.m.s. continuous sine wave signal, at the frequency to which the band

3.7. THE CISPR QUASI-PEAK DETECTOR RECEIVER

31

CISPR QUASI PEAK RECEIVER CHARACTERISTICS

QUANTITY

BAND A

BAND B BAND C BAND D

Band lower limit

9 kHz

150 kHz

30 MHz

300 MHZ

Band upper limit

150 kHz

30 MHz

300 MHz

1000 MHz

Filter bandwidth

200 Hz

9 kHz

120 kHz

120 kHz

Charging time constant 45 ms

Discharging time const 500 ms

1 ms
160 ms

1 ms
550 ms

1 ms
550 ms

Mechanical time const

160 ms

160 ms

100 ms

100ms

Standard impulse
Impulse rp rate

13.5 uVs
25 Hz

0.316 uVs
100 Hz

0.044 uVs
100 Hz

0.044 uVs
100 Hz

Equivalent cw

1 mVr.m.s

1 mVr.m.s

1 mVr.m.s

1 mVr.m.s

Overload factor (a.c.)

Overload factor (d.c.)

24 dB
6 dB

30 dB
12 dB

43.5 dB
6 dB

43.5 dB
6 dB

Figure 3.11: Summary of parameters for the quasi peak detector receiver.

32

CHAPTER 3. EMC MEASUREMENT INSTRUMENTS

pass filter is currently tuned. Such an input may be produced by applying a source of 2
mV r.m.s., in series with an impedance equal to the input impedance of the receiver, to
the input terminals of that receiver.
In one sense we may think that including the last information in the specification table
makes the quasi-peak detector receiver over-specified . If one analyses the circuit described
earlier, and assumes that at the band centre the filter has unity gain, and the indicating
meter has unity gain, then it is found that the circuit does have the property that impulses
of the size specified, oered at the rate specified, do produce the same output as does a 1
mV sine wave. If this view is taken, we might regard the entries in the table as providing
a description of performance at various impulse repetition rates rather than a component
of the specification.
The entries in the last two lines of the figure define the overload capacity required
for various portions of the receiver system. The system will only work correctly if the
impulses are not pre-clipped before they reach the detector, and if the circuits following
the detector respond linearly for signals in the range which can occur therein.
One matter not appearing in the figure but forming an important part of the definition
of the receiver is that the standard size impulses for each band are every short indeed.
It is stated that they must in each band be short enough for the power spectrum they
produce to be substantially uniform over the entire band . A simple calculation, using the
well known relation between pulse rise time ane bandwidth, will allow one to realise just
how short are the impulses in relation to the commonly applied repetition rates.
The fact that the impulses cover the entire band will allow one to appreciate why in
the above discussion we have not paid attention to just where in the band the tuneable
filter is tuned. The standard size impulses, oered at the standard rate, give the required
output no matter where in the band the band pass filter is tuned. These matters will
hopefully become clear when we examine waveforms in various sections of the receiver.

3.7.7

Output variation with impulse rate

Above we have given the standard size impulses to produce standard output when the
impulses are oered at the standard rate, which is either 25 impulse per second (Band A)
or 100 impulses per second (other bands). The question arises of how the output varies
when standard impulses are oered at other rates. For the solution of this problem, a
numerical analysis is required. The results are normally expressed in terms of how the size
of the input impulses must be increased to maintain standard output when the repetition
rate is varied from the standard rate.
The results of this calculation are shown for all bands in Figure 3.12. It may be noted
that each column has been normalised to 0 dB at the standard impulse rate of 25 per
second for Band A or 100 per second for Bands B, C and D.
Of particular interest is the fact that a value for an isolated impulse is shown. This
may at first appear as a puzzle, as impulses oered at a regular rate can be expected to
give a steady output if the rate is high enough, but a single impulse will produce a result
which is fleeting and soon gone.
The puzzle is resolved by stating that for all entries in the table the maximum value of
the output must be taken as the reading, even if the pointer of the indicating instrument is
fluctuating or exhibits but a momentary flicker. Then one may attach a reading procedure

3.7. THE CISPR QUASI-PEAK DETECTOR RECEIVER

Relative equivalent level of pulse
(dB)

Repetition frequency
(Hz)
1 000
100
60
25
20
10
5
2
1
Isolated pulse

33

Band A

Band B Bands C and D

4.51.0
4.01.0 0(reference)
3.01.0

0(reference)

+6.51.0
+4.01.0 +10.01.5
+7.51.5

+13.02.0 +20.52.0
+17.02.0 +22.52.0
+19.02.0 +23.52.0

8.01.0
0(reference)
...

+9.01.0
+14.01.5

+26.02.0
+28.52.0
+31.52.0

Figure 3.12: Response of quasi-peak detector receiver for various impulse rates.
to ones expectation of instrument behaviour for not only frequently oered impulses but
also for very infrequent impulses and for an isolated impulse.

3.7.8

Interior waveforms

The understanding of the quasi peak detector receiver is enhanced by a knowledge of the
waveforms which occur within its sections for various input conditions.
Figure 3.13 shows, in Band B, for standard size impulses of 0.315 Vs oered at a
standard rate of 100 per second, with one particular but unimportant setting of the centre
frequency of the tuneable bandpass filter, the output waveform of that filter which is ringing at its centre frequency. We see what has happened is that the filter has transformed an
impulse of indefinitely large height and indefinitely short duration (but definite amplitudeduration product) to a oscillation of definite envelope amplitude and definite duration,
the amplitude being much less than the original impulse height, and the duration being
much greater that the impulse duration.
The oscillation frequency in the waveform is characteristic of the filter centre frequency,
and not the original impulse, as that contained frequency components spread uniformly
over the full band.
During portions of the positive half cycles of the output waveform, the reservoir capacitor C increases its charge, but does not charge to the peak value of the filter output.
After the charging has taken place, there is a slow discharge, determined by the product
R2 C, until the next impulse arrives. The average charge on the capacitor is determined
by a balance between the charging and discharging process. For infrequent impulses, the
capacitor voltage remains much less than the maximum value of the envelope of the filter
output.

3.7.9

Response to isolated impulse

The response of the entire system to two infrequent impulses is shown in Figure 3.14.
The two input impulses are of standard size, and are separated by two seconds. Noting

34

CHAPTER 3. EMC MEASUREMENT INSTRUMENTS

Figure 3.13: Waveforms in quasi-peak detector receiver for standard impulse rates.

3.7. THE CISPR QUASI-PEAK DETECTOR RECEIVER

Figure 3.14: Indicator response to infrequent impulses.

35

36

CHAPTER 3. EMC MEASUREMENT INSTRUMENTS

the scales, we see that the inertia of the indicating instrument has given a rise time to
this part of the circuit which is much greater that the rise time of the envelope of the
filter output, and the total deflection of the indicator is much less than the height of the
envelope of the transient in the filter output.

3.7.10

Response to continuous wave signals

In the calculation of the response to continuous wave signals things are simple. The filter
output
will be a sin wave of the same amplitude as the input, and will have peak value
of 2 times
the r.m.s. value. For standard continuous wave input of 1 mV r.m.s. this
will be 2 mV. Despite the resistor R1 , because the resistor R2 is so much greater, the
reservoir capacitor will charge up to the peak value of the signal emerging from the filter.
The inertia of the indicating instrument will not for the steady input it receives deter
it from indicating exactly that input. Thus we may expect forthe standard continuous
wave input of 1 mV r.m.s. an indicated output of very nearly 2 mV.

3.7.11

Response to modulated continuous wave signals

One form of signal of interest to manufacturers of coded electronic labelling systems is a

carrier wave signal, which is on-o modulated with on times of about 3 ms and o times
of about 97 ms. Although this signal has a pulse repetition frequency of only 10 pulses
per second, the relatively long on time of 3 ms, compared with the charging time of the
quasi-peak detector, will ensure that the voltage on the reservoir capacitor will during the
charging period reach close to the full peak value of the filter output.
During the time between pulses, the reservoir capacitor will discharge a useful amount,
but because the discharging time of 97 ms is smaller than the time constant of its discharge,
there will be considerable residual charge. The net eect of these factors will be that the
quasi-peak detector will give a reading only a few dB at most below that of a peak detector
reading the same signal.

3.7.12

Factors regulating output

In the operation of the quasi-peak detector, one may identify six time parameters both
characteristic of the system or its input, and significant in its operation, and one time
parameter of no significance. The non-significant time parameter is the period of the
frequency to which the filter is tuned. The six significant time parameters are
The inverse of the filter bandwidth, this being related to the duration of the transient
in the envelope of the filter output when excited by a short impulse.
The charging time of the rectifier section.
The discharging time of the rectifier section.
The impulse repetition frequency when true impulses are oered, or the signal repetition frequency when pulse (but not impulse) modulated signals are oered.
The pulse signal duration when pulse (but not impulse) modulated signals are offered.

3.7. THE CISPR QUASI-PEAK DETECTOR RECEIVER

37

The inertial time constant of the indicating meter.

The combination of the non-linearity of the system and the plurality of significant
parameters makes the response of the system at times dicult to foresee clearly, but with
the aid of numerical analysis and examination of waveforms at all parts of the system, a
good understanding of why the circuit responds in particular ways to particular waveforms
can generally be gained.