Time delay measurements: Estimation of the error budget

The measurement of the time difference between two events consists in estimating, quite often with a Time Interval Counter (TIC), how much time elapses between two time markers. A time marker is defined here as the instant when a "two level" pulsed electrical signal crosses a certain threshold. In this paper we will restrict the estimation of the error budget to the case, usually encountered in a time laboratory, of the measurement of the time difference between two coherent series of regularly spaced time markers. Other uncertainty components must be taken into account when a single event is compared to a time scale. Two series of time markers, usually One Pulse Per Second (1PPS), are coherent when issued from the same clock, or from two similar low noise clocks with a small and predictable drift. In this later case the evaluation of the time delay at a given time is a little bit more complicate, but the uncertainty budget will be similar to the common clock configuration, with the statistical part taking into account also the short term noise of the clock difference instead of the noise of the measuring system alone.