Back to Westbay Tools
Click on the following bookmarks for further details of an EMC Shielding Tool:
Enclosure
Shielding Effectiveness
Waveguide below cutoff
Cavity Resonances of an
Enclosure
Twisted Pair Rejection Ratio
Circular/Square EMI
Gasket Groove Design
Aperture Shielding
Effectiveness
Enclosure Shielding Effectiveness
The Shielding Effectiveness Tool allows the attenuation of a
conductive shield to be calculated and plotted. The attenuation consists of three
components: Reflection Loss, Absorption loss and a ReReflection correction factor. All
three are calculated, along with the resultant total attenuation, and displayed on the
form shown below.
The shield material is defined in terms of its
conductivity, permeability and thickness. As an alternative to conductivity, an ohms per
square value can be entered. This can be used with the entered conductivity, in which case
the material thickness is calculated for the given ohms/square. Alternatively, the entered
thickness and ohms/square can be used to define the material conductivity. A pulldown
list of materials provides instant reference to the conductivity and permeability of
Aluminium, Brass, Chromium, Copper, Steel, Mumetal, Nickel, Silver, Tin and Zinc.
The source details are defined in terms of frequency in MHz, and
distance to the source.
The Analyse button produces a spot calculation of the Reflection Loss,
Absorption loss and ReReflection correction factor, along with the overall Shielding
Effectiveness, at the selected frequency. (Skin Depth is also calculated and displayed for
that frequency). A Tabulate button allows a table of shielding effectiveness values to be
produced over a userdefined frequency range.
The Plot command those of the above factors selected for plotting by
the check boxes on the form. A typical plotted output is shown below.
Plot of total shielding effectiveness (Curve 1) for a 0.5mm thick copper
shield, radiated with plane waves from a distant source. Note that the reflection loss
(Curve 3) falls with increasing frequency, whilst the absorption loss (Curve 2) rises.
ReReflection loss is insignificant in this example.
Waveguide below
cutoff
The effectiveness that holes in an enclosure wall transmit electromagnetic
waves into the equipment may be greatly reduced by increasing the depth of the hole. The
diagram on the waveguide form below shows a metal spindle fixed behind a hole in the
shield, and electrically connected to the shield.
The spindle acts as a waveguide, which will not support efficient
transmission below a cutoff frequency.
For a depth (d) to width (w) ratio of 3 to 1, an attenuation of 82dB is
obtained. For a width (w) of 10mm, the cutoff frequency of the spindle arrangement is
15GHz.
The cutoff frequency, Fc is given by:
Fc = 1.5 * 105 / w
MHz, where w is in mm.
The attenuation A at frequency F is given by:
A = (27.2 * d * ( 1  (
F / Fc )^{2} ) ^{0.5} ) / w dB,
where d and w are in mm.
The tool also allows for the effect of multiple apertures in the
shield, where these are less than half a wavelength apart. Thus the performance of
honeycomb vents, which are commercially available may be estimated. The reduction in
shielding caused by multiple apertures is approximated by:
Reduction = 10 * log 10
(Number of Apertures)
Thus 10 apertures would have an attenuation 10dB worse than a single
aperture, and 100 apertures 20dB worse.
Up to five sets of data are supported, and selected using the Model
Number pull down box.
Cavity
Resonances of an Enclosure
A metal enclosure can behave as a resonant cavity, with
modest size enclosures resonating in the frequency range of emc susceptibility tests. The
resonance establishes a standing wave, with the electric field peaking in the centre, and
the magnetic field peaking at the edges.
Many modes can be supported, leading to a series of peaks at increasing
frequencies. Although the calculation performed by the model is unlikely to be especially
accurate, due to the effects of the box contents, it does give an indication of the lowest
possible resonance for a particular set of dimensions.
The first sixteen resonant frequencies (of a theoretically infinite
number) are listed. Note that when two or more box dimensions are equal, modes with
different field patterns, but equal frequencies are created. These are known as degenerate
modes.
The tool requires the dimensions of a rectangular enclosure to be
entered, in mm. The Analyse button then lists the first sixteen resonances in a pull down
box, sorted in order of frequency.
Thus the lowest resonant frequency supported by a small enclosure
measuring 60mm by 70mm by 80mm would be 2.85GHz.
The expression used to calculate the series of resonances is as
follows:
Fres = 47.75 * ( ((m * Pi) / h) ^{2} + ((n * Pi) / d) ^{2} + ((p * Pi) /
w) ^{2} )^{ 0 .5} MHz
where h, d and w are the height, depth and width of the box
respectively, expressed in metres. m,n and p represent the mode integers for the
transverse electric and transverse magnetic waveguide modes TEmnp and TMmnp. Only one of
the integers m,n and p may be zero at any one time.
Twisted Pair Rejection Ratio
A twisted wire pair offers a low cost means of reducing
differential mode coupling, and balancing common mode capacitance to ground. Each twist in
the wire induces an opposite voltage to the previous twist, and for a uniform external
field, cancels the induced voltage. At low frequencies, twisted pair wire is especially
effective at reducing magnetic field induced currents.
At higher frequencies however, phase differences between adjacent twists
progressively reduce the twist effectiveness. It is this effect which is modelled by the
program. The reduction in rejection ratio can be offset at any particular frequency by
reducing the separation between twists. As with most other tools, a spot calculation and a
graph plot are both available. The only parameters to be entered are the total wire
length, the distance between twists, and the spot calculation frequency.
A series of rejection ratio plots are shown below.
Twisted pair rejection ratio for a 10 metre length of wire, for twist
separations of 0.1m (Curve 1), 0.2m (Curve 2) and 0.3m (Curve 3).
Circular/Square
EMI Gasket Groove Design
Conductive gaskets form a convenient means of ensuring
the continuity of a shield at an interface between, typically, a box and its lid or other
access panel. Flat gaskets can be used, although their compression is then dependant on
close control of the torque used to tighten the screws securing the lid. Circular or
rectangular gaskets however can be placed in machined or cast grooves, when their
compression is independant of the fastener torque, and only dependant on the groove
dimensions.
It is important that conductive EMI gaskets are placed in correctly
designed grooves, to ensure correct operation. The gasket must be deflected from its round
shape by a controlled amount, typically between 10% and 18%. Too little deflection could
result in poor emi shielding; whilst too great a deflection could cause gasket damage. The
groove depth controls deflection.
Equally the maximum gasket volume must not exceed the groove volume, or
the gasket could be extruded out of its groove, with attendant damage. All mechanical
tolerances must be combined to assess the worse case effects, which with small diameter
gaskets can produce out of limit deflection or groove overfill. Finally, if the equipment
lid or panel which is compressing the gasket is thin, and the fasteners are widely spaced,
the lid could be deflected by the gasket's deflection force. This has the result of
reducing the minimum deflection, and is discussed further below.
The Gasket Groove Design tool provides a complete analysis of the
groove design, at nominal and maximum/minimum tolerance conditions. The cross sectional
area of the groove is calculated, and compared with the gasket volume, to calculate the
groove fill. The maximum and minimum gasket deflection is also calculated. Use of the tool
will demonstrate that small gaskets, under 2mm diameter for example, require increasingly
tighter tolerances to meet the deflection requirements. Similarly, it is harder to design
a satisfactory groove for a rectangular gasket than for a circular gasket. A set of
'industry standard' diameters and tolerances for circular gaskets are provided in a
pulldown box, although any diameter can be entered.
A printed report summarises the analysis, and a summary of the groove
design is shown in a Comments section.
Lid Deflection Effect. A thin metal panel used to compress the emi
gasket, combined with large fastener spacing can result in a further reduction in the
minimum gasket deflection. This can then result in shielding degradation. The effect
increases for narrower flange width, thinner panel thickness, and less stiff materials
(lower Young's Modulus).
The contribution of lid deflection is accessed from a separate form,
opened from the main gasket analysis form. Several materials are available in a pulldown
box with their respective Young's Modulus figure.
The gasket deflection force is quoted in the nonmetric units of pounds
per linear inch of gasket material, as this is commonly used by gasket manufacturers. The
actual figure is typically 2 to 5 lbs/in for small gasket diameters (<3mm), and up to
15 lb/in for 6mm diameters. Manufacturers' data should be consulted for accurate figures,
which should be at the minimum deflection point, generally around 10%. The calculated
figure assumes constant Gasket Deflection Force, which is acceptable for small reductions
in deflection.
Aperture Shielding
Effectiveness
An opening in the wall of a shielded enclosure is a
potential source of radiated emissions, and radiated susceptibility problems. The
degradation in shielding effectiveness increases progressively as the size of the aperture
increases, until no shielding is provided when the aperture's largest dimension equals a
half wavelength. The Data Entry form is shown below; a spot calculation at a given
frequency can be made, and a graph of aperture attenuation versus frequency can be
generated. Up to five models can be plotted, and a replicate button is provided to copy
Model 1 data to the other four models.
The Aperture model requires the shield wall thickness to be thin
compared with the aperture size itself. For openings which are not round or square, the
largest dimension should be used.
The attenuation provided by an aperture is calculated from the
following equation:
Attenuation = 20 * log_{10} ( 0.5 * Wavelength / Aperture size)
dB
Multiple apertures spaced less than half a wavelength apart produce a
shielding effectiveness reduction given by:
Reduction = 10 * log_{10} ( n ) dB
where n is the number of apertures.
Use of the tool will show that small openings of the order of a few mm
will have relatively little effect at practical frequencies. Larger openings may cause
significant emc problems, and techniques such as the use of a waveguide below cutoff, or
a conductive window may have to be considered.
A typical plot of aperture attenuation versus frequency is shown below,
for five apertures of varying size. Curve 1 has a largest aperture dimension of 100mm, and
is followed by aperture sizes of 50mm, 20mm, 10mm, and finally in Curve 5, 5mm.
Note the reduced attenuation as the aperture dimension increases, and
the reduced attenuation as the frequency increases.
The Aperture Tool can also be used to plot the radiated field due to a Differential
Mode Current Loop. The details of this are entered on the DM Current Loop Tool (PCB Menu),
and this plot option can be selected by checking the Leakage Field plot box next to the
plot command on the Aperture Tool. If values are available from the Fourier analysis of a
Trapezoidal waveform, the option to perform a Fourier plot is also offered. Thus three
types of plot can be generated from the aperture tool:
* Plot of aperture shielding effectiveness, in dB's.
* Plot of the field emissions from a current loop, attenuated by the aperture, with units
of dBuV/m. The plot is continuous with frequency, and is compared with the unattenuated
field.
* Plot of the field emissions from a current loop, attenuated by the aperture, with units
of dBuV/m. The field is plotted at discrete frequencies which originate from the Fourier
analysis tool.
