An Ethernet Cable Discharge Event (CDE) test and measurement system

Abstract — A Cable Discharge Event (CDE) is an electrostatic discharge between a cable and a connector. CDEs anoccur on unshielded Ethernet based communication interfaces and inject currents into the pins directly [1-3]. The charging processes are in general understood; however, the discharge processes are complicated due to the number of pins involved and their connections to a system. Based on an understanding of the factors which determine the severity of a CDE, this article describes how to setup a variety of repeatable CDE tests and how to analyze the measurement results. Keywords — Cable Discharge Event (CDE) Test; Cable ESD; I. I NTRODUCTION Both occurrence rate and severity of a CDE needs to be considered when determining the importance of dedicated CDE tests to ensure a reliable system. The occurrence rate of a CDE depends on the type of connector used and the environment the system is used. While a USB connector A.on a laptop will receive many more ESD events than a LAN connector, the LAN connector still poses a larger risk to the system as it is usually not shielded and often used in applications that require high reliability such as back bone internet routers. In contrast to a USB connector, a LAN connector can have a spark from the connector to a pin during a CDE. In the case of USB connectors the connector shells will mate first. Provided that the shells are connected to the shield and to the system ground, most of the discharge current will flow on the connector shells. For the LAN CDE case a good understanding of the dominating processes provides a mean for developing and validating models. These models will allow computer simulation, and laboratory test setup formulation for reproducing real CDEs. This is crucial for IC and system level engineers to study and optimize the immunity of Ethernet based communications interface designs. It is well known that triboelectric charging is the culprit that generates the charge on a cable [1],[2]. This fact is especially important in Ethernet communication systems due to the long lengths of UPT cable that must be pulled through conduit, which may result into large charges. A model for describing the interaction of one twisted pair in a UTP (Unshielded Twisted Pair) cable is presented in [2] using a three body model analyzed from an electrostatic point of view. In [3] a system for discharging a cable using a relay on a test bed is presented, and experimental results are shown. In regards to the measurement alysis side of Ethernet CDE, there are few publicationsdiscussing the phenomena. Some ESD test setup have been developed to repeat ESD current transfer through Ethernet magnetic effects in [4] and [5]. The focus of this article is to present a CDE testing system that will allow for real world test conditions to be reproduced in the laboratory, thereby providing design engineers a quick and reliable method for testing new hardware designs. The test system will allow for different types of copper based Ethernetcables, and twisted pair termination strategies. Further, full control of the charge and discharge sequence of each cable line allows for all possible cases to be explored. II. I