CHAIR OF

ELECTROMAGNETIC

COMPATIBILITY

Background and problem: The theory of Schelkunoff to calculate the shielding effiency of a planar metallic wall is a standard tool for every EMC engineer to approximate the shielding effect of boxes or enclosures. The calculation is done in the frequency domain, where a harmonic excitation is assumed. Nevertheless, measurement methods and characteristics to assess the transient shielding efficiency for certain pulses are also proposed in the literature.

Task: In the scope of this work, the practicability to convert the well-known Schelkunoff theory from frequency into time domain shall be analyzed. At this, a direct approach in the time domain as well as a transform from frequency into time domain shall be investigated. This inverse Fourier transform can be done analytically or numerically. The proposed procedure shall also be tested for some typical pulse shapes of the exciting external field.

• Literature survey about the existing Schelkunoff theory
• Literature survey for transient assessment criteria of the shielding efficiency
• Development of a direct transient approach for analyzing the shielding efficienty
• Transform of the existing frequeny-domain solution into the time domain
• Test of the procedure for some standard pulses

Supervisor: Dr.-Ing. Mathias Magdowski

Background and problem: Many interference phenomena of electromagnetic compatibility like galvanic, capacitive and inductive coupling as well as the corresponding countermeasures like bonding, filtering and shielding can be more easily understood if they are practically demonstrated. Therefore, EMC demonstration units or "demo boxes" have been popular for several decades already.

Task: A new EMC demonstration unit shall be designed, developed and constructed. It shall be based on typical designs that are described in the literature. The box should be simple in design, mechanically and electrically stable, easy to transport and to use. In contrast to the typical designs in the literature, where a full-fledged spectrum analyzer is necessary for the demonstration, a much cheaper SDR (software defined radio) receiver shall be used here, that only requires a standard computer with some USB port.

• literature research about existing EMC demonstration units
• design and development of an own demonstration box
• setup and construction of the demonstration box
• bringing this box into service

Supervisor: Dr.-Ing. Mathias Magdowski

Background and problem: Reverberation chambers are commonly used to test the immunity against high intensity radiated fields. The chamber acts as a resonator with a preferably high quality factor and low losses. In the steady state, the input power equals the power loss. From the difference between the power loss in the empty chamber and the chamber loaded with the device under test (DUT), the coupled power to the device under test can be determined.

Task: Such an indirect measurement shall be done in different frequency ranges with diverse DUTs in the three mode-stirred chambers of the chair for electromagnetic compatibility. For simplicity, plain monopole antenna with one main resonance shall be used as a DUT. The resonant frequency and bandwidth of the DUT shall be determined from the frequency dependence of the coupled power to the DUT.

The experimental results shall be validated by a direct measurement of coupled power to the DUT. The discrepancies between both measurement shall be discussed. Also the uncertainty of the indirect measurement as well as its reasons shall be analyzed.

Supervisor: Dr.-Ing. Mathias Magdowski

Background and problem: The field in reverberation chambers can be described statistically. This description covers the distribution of the field quantities at one position as well as the spatial correlation between nearby field points. Usually, the field are assumed to be circular, which means that the real and imaginary parts of the complex phasors of the field components are independent of each other, but follow the same distribution. From this follows that the field is statistically homogenous, isotropic, unpolarized and incoherent. Based on this assumptions, e.g. also the maximum values of the field components and therefore the failure probability of an equipment under test can be determined.

In practice however, the field will always feature a certain ellipticity, i.e. a difference between the real and imaginary parts of the complex field components. Measurement are necessary to determine the actual field properties in mode-stirred chambers. Such measurements have only been done with linear polarized antennas up to now.

Task: The aim of this project is to measure the complex scattering parameters between a linear and a circular polarized antenna as well as between two circular polarized antennas. A vector network analyzer is available for this measurement. As linear polarized antennas, different logarithmic-periodic dipole and horn antennas are provided. A helix antenna is available as a circular polarized antenna. A second helix antenna has to be build according to this prototype. The measurement of the scattering parameters has to be done over a wide frequency range for different stirrer positions and has to analyzed statistically.

Supervisor: Dr.-Ing. Mathias Magdowski

Background and problem: Cables are important coupling paths of external electromagnetic fields into connected devices and systems. In practice, not only single cables but cables harnesses occur that can be regarded as transmission line networks. External fields can often be approximated as plane waves, at least in the far field region.

The simulation of the plane wave coupling to transmission line networks is quite well understood in the frequency domain. Nevertheless, when the loads at the terminals of the network feature a non-linear behavior, e.g. as for a diode, the calculation has to be done in time domain. Such calculation has already been done for a single cable, where the exciting field was incorporated as several distributed sources along the line.

Task: The task of this project is do adopt this approach for a transmission line network by taking into account the interaction between the individual lines. For simplicity, the lines can be assumed to be straight, uniform and of low loss. For linear loads and a certain pulsed excitation, the time response of the coupled voltage or current should be validated against a frequency-domain solution with a following inverse Fourier transform. Then also the time response of the coupled voltage or current for a non-linear load shall be calculated.

Supervisor: Dr.-Ing. Mathias Magdowski

Background and problem: The field in electrically large and complex shaped resonators (e.g. car bodies, aircraft fuselages, ...) can in principle be described deterministically. Anyhow, such a description is of little value, as a small change in frequency, in the spatial position or in the electromagnetic boundary conditions may lead to a completely different field pattern. Therefore, a statistical field description is much more suitable that can also be experimentally reproduced in reverberation chambers. If a device under test is placed in such a field, also the coupling has to be described statistically. For this, several methods exist, as the Random Coupling Model or the Plane Wave Integral Representation.

Task: The aim of the project is to solve a given coupling problem with both methods and to compare both procedures (e.g. necessary parameters, computational effort, accuracy, ...). As a coupling problem, the field coupling to a single wire transmission line above a conducting ground plane shall be analyzed. For this problem, experimental results as well as several analytical and numerical results based on the Plane Wave Integral Representation exist at the chair for EMC, so that only a solution via the Random Coupling Model has to be found for comparison.

The solution to be analyzed is e.g. the coupled voltage (or current) at one line end as a complex phasor. This phasor can be characterized by its real and imaginary part, its magnitude and phase or its squared magnitude, which is proportional to the power. From these characteristics, the frequency-dependent average, minimum, maximum or standard deviation can be calculated. Also the probability density function, cumulative distribution function or the general statistical moments are of interest.

Supervisor: Dr.-Ing. Mathias Magdowski