Page 357 - Fiber Bragg Gratings

334 Chapter 7 Chirped Fiber Bragg Gratings

a gap of 10.25 A Bragg. The group delay is more sensitive but is only affected
when light crosses the gap. This means that light launched from the long-
wavelength end experiences a change in the group delay ripple on the
short-wavelength side of the gap, and vice versa when traveling from the
short-wavelength end. This has implications for SSCGs made with more
than two SCGs. The grating nearest the launch end remains essentially
unaffected, while the group delay ripple of subsequent gratings deterio-
rates.
Figure 7.19 shows the effect of the join on the group delay. At the
join, there is a localized discontinuity, which becomes narrower as the
gap gets larger, and almost disappears. Simulations have shown that
large gaps (5 mm) tend to smooth out the effect of the join, but the delay
through the gap does introduce a step change in the group delay on either
side of the join.
We see a relative increase in the noise from the long-wavelength end
(launch end). In comparison, a 500-mm-long SSCG grating with random
stitching errors at every 100 mm is shown in Fig. 7.20 [46]. Here, the
stitching produces small spikes at the join, and also a general increase
in the noise at the short-wavelength end. These are believed to be due to

Figure 7.19: The group delay at the join of the gratings shown in Fig. 7.18
for two values of the gap: 0.25 (A) and 10.5 (B) times the Bragg wavelength. The
delay spike is localized to the join and is not apparent from the long-wavelength
end of the grating. In this simulation, the light enters the long-wavelength end
of the grating. The effect is reversed if the light is launched from the short-
wavelength end.