Improving the Accuracy of Time-Difference Measurement by Reducing the Impact of Baseline Shift

Yang Zhang,Zhaohui Li
DOI: https://doi.org/10.1109/TIM.2015.2437635
2015-01-01
Abstract:Many time-difference measurement devices such as time-to-digital converters (TDCs) are designed on the basis of threshold or zero cross detection technology. However, baseline shift, which means that the detected signals may have a shift of amplitude of the whole signal, is common in electronic circuits and has a great influence on the accuracy and stability of the time-difference measured with conventional posedge or negedge detection methods. In general, the baseline shifts can be classified into two cases which are parallel and oblique baseline shifts according to their different origins. Parallel baseline shift means that the baselines of two detected signals are parallel but in different levels, while oblique baseline shift is that the baselines of two detected signals are not parallel and the baselines have a nearly linear relationship with the time in the duration of signals. In this paper, detection methods using multiple edges, namely, double-edge and triple-edge methods, are proposed to reduce the impact of relative baseline shift between two signals for high-precision time-difference measurement because they require more information of the signal waveforms than conventional methods. The simulation results show that the proposed methods have a better performance than conventional posedge or negedge method. The experiment results based on a TDC chip, i.e., TDC-GP21, also demonstrate the validity of proposed methods when baseline shifts exist.
What problem does this paper attempt to address?