Optimization of dual-mixer time-difference multiplier

The dual-mixer time-difference (DMTD) multiplication technique used in frequency stability measurement of precision frequency sources shows potential for achieving a frequency stability measurement floor as low as a few parts in 10¹⁵ at 1 s in terms of Allan deviation (ADEV) at 5 MHz at ambient temperature. The paper discusses the optimization of the DMTD multiplier using the frequency stability floor as a criterion for its efficiency. The instability sources of concern are all the noise sources as well as non linear effects and deterministic effects, such as phase noise in the common oscillator, noise in the mixers and amplifiers, intermodulation products including the power-line pickup, electrical asymmetry (impairs the common-mode rejection of the phase and amplitude variations), phase dependence on environmental fluctuations, disturbing electromagnetic fields, and inappropriate construction (connectors, cabling, shielding, grounding). The optimization is demonstrated on the second version of DMTD multiplier designed at the Institute of Radio Engineering and Electronics (IREE). The IREE system has the "classical" DMTD structure, i.e. with a time-interval counter measuring the time difference between the zero crossings of the two beat-note signals. The modular construction allows modifications of the DMTD critical elements. The common oscillator as well as the time-interval counter are located outside the mixer-amplifier unit. The system can be operated at 5 and 10 MHz with the beat frequencies of 5 and 10 Hz, respectively, giving the multiplication factor of 10⁶. The frequency stability measurement floor achieved through the optimization shows ADEV(tau = 1 s) = 6.9times10^-15 at 5 MHz. At the basic sampling interval, tau₀ = 0.2 s, the time deviation TDEV is 3.5 fs with the TDEV floor of 2 fs. This outstanding performance makes it possible to measure the frequency stability of future 5MHz quartz crystal oscillators of a f