Personnel Electrostatic Discharge: Impulse Waveforms Resulting from ESD of Humans Through Metallic-Mobile Furnishings Intervening in the Discharge Path

This study represents an extension of an effort that was originated by King and Reynolds in 1980, and presented to the 1981 IEEE EMC Symposium (1) held at Boulder, CO. In the 1981 presentation, ESD impulse waveforms were re­ ported as measured from two conditions: Human-direct; and, human through small hand-held metallic objects that were intervening in the discharge path. Using a co m ­ patible measurement approach, this effort expands the study to be inclusive of metallic-mobile furnishings found typically in the office and/or data processi.ng environment, such as desk chairs and push-carts. Cursory sample measurements were additionally taken to illustrate impulse waveform variability at selected amplitudes, describing different effects due to different Electrostatic Discharge locations on the surfaces of a desk chair. Other sample measurements provide indication of other impulse results if the desk chair were to be replaced by a stool-type "lab chair". As in the previous publication, the conceptual goal of this effort was to develop information derived from actual ESD events that would facilitate the design of an ESD test generator which could produce impulse waveforms in reasonable but realistic simulation of actual ESD events for these ESD "furnishings" event conditions. Retrospective Overview As reported in the 1981 publication, the impulse charac­ teristics of ESD dynamically varied as the initializing (charge) amplitude was varied. The dramatic alterations encountered in the waveshapes as the initializing level was incrementally increased confirmed the hypothesis that waveform measurements derived from incrementally varied amplitudes were required to fully characterize the ESD event continuum. Generally, it was found that the impulse waveforms derived from lower-level initial amplitudes exhibited ultra-fast risetimes between 200 picoseconds (the limit of the measurement capability) and 500 picoseconds, while the impulse waveforms derived from higher static levels exhibited risetimes in the approximate range between one nanosecond and several tens of nanoseconds. The waveform components providing the ultra-fast risetimes were also found to be confined to exceptionally small pulse widths, typically in the area of a few hundred picoseconds, developing intense currents between the general range of a few tens of Amperes to over 100 Amperes. The more-conventional ionization-based impulses developing risetimes of one nanosecond (or more) exhibited peak currents up to a few tens of Amperes with base widths up to approximately 500 nanoseconds, although typical values were usually less than 200 nano­ seconds. The development o f multiple impulses within the framework of what was considered a 'single' ESD event was investigated in the previous effort, as was the duty cycle (or, periodicity) among events within the time envelope of 'one' event. Although the suggestion that ESD events were encountered with risetimes as fast as 200 picoseconds was recognized as unconventional (if not controversial) by the authors at the time of publication, another presentation by Carruth, et al, (2) at the same symposium hypothesized that dielectric discharges may exhibit risetimes faster than 100 picoseconds, based on statements made by Leung, the presentor. Given the back­ ground of the study effort provided above, measurements were taken on the ESD dynamic characteristics of impulses DISCHARGE PATH David Reynolds Electromagnetic Compatibility Engineer Digital Equipment Corporation 301 Rockrimmon Blvd. South Colorado Springs, Colorado 80919 developed from human interaction with mobile-metallic furnishings, based on the measurement techniques pre­ viously used, and with the ESD initializing amplitude incrementally ascending to determine the impulse wave­ form alterations attendant to the amplitudes. Measurement Approach As in the previous study, the conceptual measurement goal was to determine the ESD waveforms and waveform characteristics for the 'source' ESD events in as low an impedance measurement system as was considered practical. The foundation for this approach is that if the 'source' characteristics were known, and adequately simulated in an identical measurement-evaluation condition, then an eventual 'system-in-test' would respond to the simulated impulses in an equal manner compared against the 'actual' ESD occurrence (through furnishings). Reference is made to the 1981 publication (1) for the details of the development of the measurement system and approach, to enhance brevity here, although the following is provided to aid understanding of this effort. The essential components of the measurement method consisted of: a) A discharge 'load', of known impedance characertisties; b) An exceptionally wide bandwidth oscilloscope system; and, c) A reference plane arrangement to 'capture' the distributive impedances from the human and furnishing under evaluation..The impedance characteristic of the 'load' was designed to partially compensate for higher frequency roll-off in the oscilloscope system, resulting in an enhanced-bandwidth measuring system. Measurement 'Discharge' Load: The discharge 'load' consisted of a circular array of nine, 1.8 Ohm carbon-composition resistors, providing an initial impedance of 0.2 Ohms. The array was matched to the 50 Ohm input of the oscilloscope, through which a coupling loss of 6dB was experienced. With the 6dB loss, the load yielded an equivalent response of 1.0 Ampere per 0.1 Volt indicated on the oscilloscope. An insertion loss profile of the load indicated a 'flat' response within 4dB from 0 1.0GHz, elevating in impedance by approximately 8dB between 1.0GHz and 1.5GHz. This impedance increase compensated for the 'scope roll-off, approximately. Figure A-l of Appendix I illustrates the insertion loss profile of the load. Oscilloscope System: The Oscilloscope System consisted of a Tektronix 7104 main frame with a Type 7A29 Vertical Amplifier and a Type 7B15 (or Type 7B10) Time Base. The system provided an instantaneous analogue bandwidth of 1.0GHz, a write speed of 200 picoseconds in the amplitude ranges used, with the bandwidth rolling-off by 8dB at 1.5GHz. The amplitude response of the oscilloscope actually used for this effort is provided by Figure A-2 of Appendix I. The frequency response deviation curve for the measurement system as utilized may be gained by combining the load response curve and the response curve of the oscilloscope system, as provided in the Appendix. The approximate net system response deviation is provided By Figure 1. Figure 1. Approximate Net Response Deviation of Measurement System.