Microwave-envelope soliton threshold powers and soliton numbers.

Abstract

The peak power of 5.8-GHz backward-volume-wave pulses propagated in a 7.2-\ensuremath{\mu}m-thick yttrium iron garnet film was measured as a function of input pulse width ${\mathit{T}}_{0}$ from 5.1 to 44 ns and input power from 0.05 to 3.5 W in a two-stripline microwave transducer structure. A threshold input power ${\mathit{P}}_{\mathrm{th}}$ for the onset of a nonlinear response was determined as a function of ${\mathit{T}}_{0}$. The dependence of ${\mathit{P}}_{\mathrm{th}}$ on ${\mathit{T}}_{0}$ followed a clear pattern consistent with theoretical considerations for soliton formation. Plots of ${\mathit{P}}_{\mathrm{th}}$ vs 1/${\mathit{T}}_{0}^{2}$ showed two regions where ${\mathit{P}}_{\mathrm{th}}$ scaled linearly with respect to 1/${\mathit{T}}_{0}^{2}$. The slopes of these two linear regions were in ratio 1:9, as expected from theory. A soliton-number parameter n calculated from the ${\mathit{P}}_{\mathrm{th}}$ data as a function of ${\mathit{T}}_{0}$ showed plateaus at n=1, 2, and 3. A one-to-one correlation was found between these calculated values of n and the number of peaks in output pulse profiles. These results demonstrate a simple connection between input pulse width, nonlinear power threshold, and soliton number.