(a) Pulse sequence to measure electrically detected spin echoes. For signal modulation, we alternately apply the spin-echo pulse sequence with the phase of the last π/2 pulse set to (+x) and with its phase set to (−x). This cycle is repeated N cycle times. The current transients (solid line) after the mw pulses consist of a spin-independent non-resonant part (dashed line) and a spin-dependent resonant part. After the (−x) spin-echo pulse sequence, the resonant contribution to the current transient is inverted when compared to the current transient after the (+x) pulse sequence. The shaded area indicates the boxcar integration interval Δt. (b) Calculated response of the lock-in detection scheme for different boxcar integration intervals Δt scaled by the indicated factors. (c) Bandwidth calculated for different numbers of cycles N cycle.
(a) Integrated charge ΔQ as a function of τ2 for τ1 = 300 ns measured with phase modulation at f mod = 100 Hz. The data points with the phase of the last π/2 pulse set to (+x) (upper trace) and (−x) (lower trace) are shown separately. (b) Echo trace obtained by subtracting the two echo traces (+x) and (−x). For comparison, the echo traces (+x) and (−x) after subtraction of the background taken as the smoothed average of the two traces in (a) are shown as well. (c) Signal-to-noise ratio of an electrically detected spin echo as a function of the modulation frequency f mod. (d) Sketch of the non-resonant (dashed lines) and resonant current transients (solid lines) with Fourier components at even multiples and odd mutiples of f mod, respectively.
Summary of the contributions of different parts of the measurement setup to the noise floor at f mod = 111 Hz. The different contributions to the noise level are assumed to be independent, so that the squares of their standard deviations can be added to calculate the overall noise level.
Article metrics loading...
Full text loading...