Noise induced time interval measurement biases

Time jitter in the output fronts of hysteretic triggers is not the only effect of additive noise. An average leading time bias error can be observed when large bandwidth noise is added to a poor slope signal. Time jitter is shown to be a possible limitation to the accuracy of high precision time interval measurements, as the average switching time is always before the zero crossing of the ideal noiseless analog signal. The resulting bias in the start-stop type of measurements, in which a noisy signal is compared to a reference, is shown to depend on S/N ratio and on noise bandwidth. An experimental evaluation of this effect is reported, based on time interval measurements in which noise is added to a sinewave signal