Measurements and modeling of EMC, applied to cabling and wiring

Abstract

Listen

Translate article

A myriad of cables transport power and communication signals in larger buildings and installations. Cross talk between cables and connected equipment is a major concern. Regulations do exist, but often show to be insufficient to avoid undesirable coupling. The current thesis research addresses this problem and provides a tested model for the interference coupling in buildings, in particular those caused by lightning. The model should also give an insight in the reliability of cabling and wiring, even when not all details of the installation are known. This was the goal as proposed in the IOP-EMVT project 'Optimal cabling in buildings and installations qua EMC'. A newly built pharmaceutical plant acted as main test object. In the measurements, currents of 0.3 kA were injected in the lightning protection grid on the roof. Inside the building, 100 m long test cables followed the path of other installation cables on the ladders and trays. The measured current and voltage are typical for the other cables. A simplified model of the installation included most designed current paths. It was implemented in method-of-moments program FEKO. Measurements and model agreed that the roof steel skeleton carried about 80 % of the current and the intended lightning conductors 20 %. A nearby, non-intended conductor (an air duct) had to be included in the model to obtain acceptable agreement between the calculated current through a cable support and the measured one. For three types of cables, the measured voltages agreed with the currents when combined with the transfer impedance measured in the laboratory. The agreement allows extrapolating the model to real lightning. This has been done in two steps, the first and simple takes the cable transfer impedances into account; the second and more complicated also includes travel time and resonances in the installation. The differences between both are limited for the Profibus fieldbus cable and not for the 2-lead cable with steel armor. The transfer impedance of the cables showed the advantages of armored cables even inside buildings. Additional interconnects to ground constructions cause a reduction of the lightning current inside a structure. They reduce the excitation of internal building resonances and shift the resonance frequencies upwards. Unrealistic artifacts in model results should be avoided by including a sufficient number of interconnects. Other shorter experiments are presented. For example: measurements and calculations on the lightning safety of an electronic lamp driver have been carried out on request of Philips Lighting. Based on the knowledge developed, an effective remedy against unacceptably large damage could be given. Simple configurations serve as direct test case for the models, such as the current distribution over a set of two 70 m long horizontal grounding electrodes. We compared measurements and the FEKO model, with simple analytical expressions. The interesting frequency range is up to 1 MHz, of relevance for lightning and conducted interference in switched mode power supplies.