We present a new adaptive subsample Explicit time delay estimator(ETDE) that is well suitable for determining and tracking the large-range time-varying delay between two noisy bandlimited signals only using low-order interpolators. Unlike conventional ETDE where a Fractional delay filter(FDF) is used, the new ETDE is based on a novel on-line interpolator which has the ability to optimally synthesize any continuously variable nonintegral sample delay. The theoretical derivations for our proposed on-line interpolator and the new ETDE algorithm are provided. Simulations are presented to evaluate the performance of the new estimator and to compare the performance with other estimators. Simulation results show that the new ETDE significantly outperforms FDF-based ETDE when the delay is large or varies in a large range,and the same filter orders are used. We present a new adaptive subsample Explicit time delay estimator (ETDE) that is well suitable for determining and tracking the large-range time-varying delay between two noisy bandlimited signals only using low-order interpolators. FDF) is used, the new ETDE is based on a novel on-line interpolator which has the ability to optimally synthesize any continuously variable nonintegral sample delay. The theoretical derivations for our proposed on-line interpolator and the new ETDE algorithm are provided. are presented to evaluate the performance of the new estimator and to compare the performance with other estimators. Simulation results show that the new ETDE significantly outperforms FDF-based ETDE when the delay is large or varies in a large range, and the same filter orders are used.