Analysis of errors in transient disturbance measurements using high-pass probes

Transient disturbances conducted on the mains are increasingly recognized as important events. Time-domain treatment is essential for the analysis of these transients, which are distinct from other disturbances (cw), where a simpler frequency-domain technique can usually be applied. Transients are measured using a wide variety of devices with a known transfer function (voltage probes, current probes, baluns) to couple the disturbed conductor to the measuring apparatus. All these devices introduce errors due to their transfer function. Most measuring devices are shown to be accurately represented by a single analytical model consisting of a first-order or second-order high-pass network, thus permitting accurate analytical and numerical determination of the output waveform for the more interesting input waveforms, i.e., unidirectional transients (double exponentials) or ringing transients (damped sines and cosines). The distortion at the output results in errors of the more common parameters used to specify the disturbances, namely, peak value, average rate of rise, integral of the squared voltage, impulse strength, and undershoot. These parameters are computed and shown in graphical form, permitting evaluation of the errors simply on the basis of the knowledge of the cut-off frequency of the probing network and of the characteristics of the input transient. Simplified asymptotic treatment is also offered for the determination of the output waveshape for early times and late times. The early-time error is shown to be proportional to the integral of the input signal, and the late-time output waveshape corresponds to the impulse response of the probing network. The above results permit a simple treatment of complex waveshapes. Experimental verification confirms the analytical and numerical predictions. © 1990 IEEE