Fig. 5: Histogram analysis of Mach–Zehnder interferometer signals.
The relative population of rubidium atoms in the excited state after Mach–Zehnder pulse sequences: (left-hand side) with respect to ϕlaser and the corresponding histograms (right-hand side). For both (a) T = 0.5 ms, and (b) T = 10 ms, N+/Ntot was determined by fitting the images to Gaussian profiles for each experimental run and the fitting errors, shown for each N+/Ntot, were propagated to the corresponding histogram errors using a Monte Carlo simulation. The histograms were fit with a PDF, described by Eq. (9), corresponding to a double-peaked distribution with amplitude A and offset P0 and broadened into a nearly flat-top distribution by Gaussian noise with amplitude σP (red curves). Although the characteristic sinusoidal dependence of N+/Ntot with respect to ϕlaser is not observable for the T = 10 ms data, a reasonable fit of the histogram with the double-peaked PDF, giving C = 0.24 and SNR = 2.4, signifies that phase-noise is a major cause of the loss of AI visibility as T is increased to 10 ms for the CAL single-species 87Rb atom interferometer.
