Page 99 - Tunable Lasers Handbook
P. 99

80      Charles Freed

                   coupling loss). In a small CO,  laser with L = 50 cm and r,. = 5%  Q,  is of the order
                   of  107; thus for a typical power output of  1 to  10 W (which is easily obtainable
                   with a small TEMOoq mode CO, laser) the quantum-phase-noise-limited linewidth
                   is less than  10-6Hz.  Note that-10-6HHz  represents less than  1 part in  1019 of the
                   output frequency (vo -3  x 1013 Hz) of a CO,  laser. This inherent spectral purity of
                   CO,  lasers can be  explained as follows: The linewidth AV is inversely propor-
                   tional  to  the  product of  Po  and  Qf,  and  the  combination of  high  Po  and  high
                   Q,  can be  simultaneously achieved with relative ease even in a small CO,  laser
                   oscillator. Oscillators in  the  radio-frequency (rf )  and  microwave domain have
                   either high Po or high Q,  but not both together in a single device.
                      Laser  stabilities are most  frequently  measured in  the laboratory from  the
                   results of heterodyne experiments with two lasers. Laser stabilities can be deter-
                   mined  by  either  frequency-domain  (Fourier  spectrum) or  time-domain  (Allan
                   variance) analysis  of  the beat-note  spectra of  the  laser pairs. To  establish the
                   spectral purity we can heterodyne two CO,  lasers of  equal high quality so that
                   the resulting beat-note spectrum can be apportioned equally to each laser. Two
                   problems arise, however, in trying to measure the Schawlow-Townes  linewidth
                   of high-quality CO,  lasers. The first of these problems is instrumental: The sra-
                   bility of  the  available instrumentation itself generally cannot reliably measure
                   spectral purities of  10-6 Hz or better.
                      The origin of the second problem is that for well-designed CO,  lasers [56]
                   the  so-called  technical noise  sources dominate over  the  quantum-phase-noise-
                   limited Schawlow-Townes  linewidth [57]. Examples of  technical noise sources
                   are acoustic and  seismic vibrations, and power-supply ripple and noise. These
                   sources can cause frequency instabilities by perturbing the effective cavity reso-
                   nance via the sum of fractional changes in the refractive index n and the optical-
                   cavity length L:
                                           Av=v (%  tL)
                                                  -+-


                   As  an  example,  a change of  only   A  (about  1/1000 of  the  diameter  of  a
                   hydrogen atom) in a 50-cm-long CO,  laser cavity will cause a frequency shift of
                   approximately  6  Hz.  4  6-Hz  variation  in  the  approximately  3 x lOI3  Hz  fre-
                   quency of a CO,  laser corresponds to a fractional instability of 2 x
                      Figure 6 shows the real-time power spectrum of the beat signal between two
                   free-running lasers that were designed and built at Lincoln Laboratory [56]. The
                   spectral width of Fig. 6 implies a frequency stability at least as good as 2 x
                   The discrete modulation sidebands in the figure were primarily due to ac-power-
                   line frequency harmonics, cooling fan noise, and slow frequency drift; however,
                   each  spectral line was  generally within  the  10-Hz resolution bandwidth of  the
                   spectrum analyzer. The measurement of the spectral width was limited to a 10-Hz
                   resolution by the 0.1-sec observation time that was set by the instrumentation, not
                   by the laser stability itself.
   94   95   96   97   98   99   100   101   102   103   104