Combined high power and high frequency operation of InGaAsP/InP lasers at 1.3 μm

Abstract

al.,' who found that for the specific case of u)/h = 0.101 and t / h = 0.011, the microstrip with a circular-edged strip will have 15% less conductor loss than the microstrip with rectangular edges. For the same configuration, the present technique predicts a 10.7% decrease in loss for a circular cross-section. Conclusion:Microstrip conductor loss can be calculated quickly by a perturbation method which includes the effect of strip edge shape. The work of Lewin and Vainshtein has been extended here to implement the technique on microstrip. The required stopping points in the loss integrations have been found here for a variety of trapezoidal strip edges. Loss calculations have been done for these and other edges, and it is apparent that the edge shape has a significant effect on the conductor loss. Although results here tend to overestimate the effect owing to breakdown of the approximations, conductor loss due to edge shape increases as the edge shape gets sharper and the strip gets narrower or thicker.

The gain and carrier density in semiconductor lasers under steady-state and transient conditions

“…frO'(k) andf(P)(k) are implicitly dependent on the optical field and injection current in the active region. To make the process self-consistent we use the very different time scales for population relaxation (r,,s) and carrier intraband scattering (re, 71, -1 0 -l~ s) to solve for carrier density in the presence of injection current and optical power [(24) and (25)]. Equations (3)-(5) can be solved directly for the perturbed distributions f,(k) and fh(k) in a manner similar to that used in a two-level system 1221…”

“…= 0.35, and ai = 20 cm-' for a InGaAsP-InP laser with mirror reflectivities R I = R2 = 0.3 and cavity length L = 300 pm. P / P s was estimated to be less than 0.2 for a InGaAsP-InP laser without facet coating up to the highest power 50 mW per facet [25]. For lasers with-even higher output power (159, (26)-(28) are not valid and one has to employ strong-signal theory to more accurately include the effect of gain saturation.…”

“…It is interesting to compare the above results with those of the linear theory when the gain dependence on carrier density is linear [(I)] even above lasing threshold. The corresponding expressions can be easily obtained by solving the rate equations (24) and (25) with GY = 0. The photon density in the linear theory is ductor lasers are oscillators only partially driven by the spontaneous emission.…”

“…We employed the time-dependent rate equations (24) and (25) to study the gain and carrier density dependence in the transient regime. The rate equations were solved numerically by assuming a constant injection current switched on at t = 0.…”