Weiner and Silber : Pulse Sharpening Effects 1473 50 01 4 Ferrite Cable Network Bias Supply Ferrite

may be described using complex shock wave analysis [ I . Fig. 1. Energy contained in rise time portion of pulse is used to reverse Recently, an appro.ximate but useful model describing this spins in ferrite transmission line. Magnetic field profile (dashed curve) phenomena was reported 121. Although some experimental in spin reversal region, as it emerges from output of line, represents • residual, sharpened rise time. work also was reported, there was insufficient data for a meaningful comparison between model and experiment. In this a ferrite transmission line which is uniformly magnetized in paper the pulse rise time emerging from the output of a ferrite the direction transverse to the direction of propagation (Fig. I). coaxial line was obtained as a function of voltage, line length, A transmission line without ferrite, with impedance Z0, is and magnetic field bias, and the results were compared with connected to the input terminals of the ferrite line. A pulse the model. Results show that the model is essentially correct, with risetime TR is incident upon the ferrite line. The polarity although a precise comparison is difficult because of the of the magnetic field of the pulse is opposite to that of the numerous factors (several of which are not well understood) magnetization. As a consequence. a spin reversal process will which influence the pulse sharpening effect. be initiated, and the pulse will see a large RF impedance Zf The motivation for this work is the need for pulseis which consisting of an inductance as well as a resistive component simultaneously satisfy pulser requirements for fast rise time caused by dissipation in the ferrite. The ferrite line will not (-l.0 ns) and high pulse repetition rates (Z20 klz) at kilovolt continually appear as a large impedance, however. Eventually, levels (15.0kV). Switches now available do not simultathe initial portion of the ferrite line will reach saturation. neously satisfy these requirements. One possible solution is When this happens the large impedance will suddenly decrease the use of the ferrite transmission line in combination with a to the saturated impedance Z, which by design is chosen equal slower rise time switch, for example, a thyratron. There are to Z0 the input impedance. As shown in Fig. 1,this process disadvantages,however,and these are added circuit complexity, continues, so that a "spin saturation front" propagates along >bulk, as well as lowered circuit efficiency caused by pulse the length of the ferrite. The incident signal will be both 0L. dissipation and bias current. Nevertheless, the ferrite pulse reflected from and transmitted beyond the front. The velocity C sharpener has potential in an area where there are few technoof this front will increase as the pulse amplitude is increased. C logical alternatives. The ferrite line is designed such that when the front reaches