Practical aspects of a pulse generator calibration

Abstract

Measuring receivers are used for measurement of radio disturbance in the frequency range typically 9 kHz to 18 GHz. Such receivers can be either electromagnetic interference (EMI) receivers or spectrum analyzers with the quasi-peak (QP) detector [1], [2]. Requirements for measuring receivers are discussed in European CISPR documents (Comité International Spécial des Perturbations Rádioélectriques) and US standards ANSI 63.2 (QP parts derived from CISPR). Although novel techniques are being evaluated [3], the traditional way of checking compliance of the receiver with requirements of these standards is using a calibrated pulse generator [4]. The calibration of pulse generators is discussed in the standard EN 55016-1-1 [5], which is the harmonized version of the international standard IEC/CISPR 16-1-1 (currently Ed. 4 [6]). In the standard [5], however, only a very brief description of the methods is given and technical details are hidden. The measurement uncertainty of the pulse generator characterization is not discussed in the standards, however, it was discussed in several previous works [7], [8]. This paper aims to provide a more thorough description of particular calibration methods together with practical hints which may be useful for students, calibration engineers and practitioners. Pulse generators According to [5], a pulse generator is an instrument capable of generating time-domain rectangular pulses, or a pulse-modulated RF signal. Rectangular pulses are typically used for lower frequencies (bands A/B), pulse-modulated RF signals for higher frequencies (bands C/D) because of the risk of receiver damage due to high peak voltages. The base-band pulse generators usually comprise of an energy-storage device (electrostatic, magnetic field) and a switch which discharges a fraction or all of the energy into a load. The pulse-modulated RF generator uses a harmonic signal with a pulse envelope. The spectrum is similar to a rectangular pulse (upconverted to the carrier frequency f c), maximum of the spectrum is at f c. The spectrum is uniform in a given bandwidth, which implies that pulses with longer duration can be used with lower amplitudes compared to base-band pulse generators (lower risk of Biography Martin Hudlička (S'04-M'08-SM'15) received the Ing. (M.Sc.

Evaluation of Spectral Estimation Parameters for Direct Sampling FFT-Based Measuring Receivers

“…Reference Signal 3 -Rectangular pulse: The last reference signal consists of a periodic rectangular pulse that emulates the one used for standard calibration of the response to pulses of the receiver [24]. The said pulses have a short duration and fast rise time, producing an approximately flat frequency spectrum in the corresponding frequency range.…”

Metrological Characterization of EMI Receivers

“…Historically, there has been confusion about the terminology used with the pulse generator characterization, see, e.g. [17]. One can find terms spectrum amplitude, impulse area, spectral intensity or spectral density in the literature, whereas the most common is the use of spectrum amplitude (ideal rectangular pulse) and impulse area (general pulse shape).…”

“…There are several methods for the characterization of nanosecond pulse generators and pulse-modulated RF generators given in [3] Annex C. However, only a very brief description of the methods is given and technical details are hidden. In [17], the main calibration methods of base-band pulse generators were compared together with the achievable measurement uncertainty (i.e. the Fourier transform of a time-domain pulse waveform, the intermediate-frequency measurement method, the measurement of pulse amplitude and duration and the measurement of one spectrum line amplitude).…”

“…[19]. The impulse area of a pulse generator determined by one of the methods discussed in [17] is then used as a reference value for the calibration of pulse-modulated CW signals. In the past, there were various nanosecond pulse generators available in the market (e.g.…”