Calibration of channel mismatch in time-interleaved real-time digital oscilloscopes

Abstract

A novel method is proposed for the calibration of channel mismatch in a time-interleaved real-time digital oscilloscope (RTDO). A simple simultaneous equation is derived from the Fourier transform of the time-interleaved signals. Thus, it only requires a transfer function of time-interleaved ADCs (TIADCs), while most of previous works have employed additional filters. A measurement method for the transfer function of a commercial RTDO is also proposed. The accuracy of the calibration method is determined by the noise produced after the interleaving process. To validate the proposed method, we measure two-tone signals using a commercial RTDO. The calibrated results clearly show signals at spurious frequencies are substantially reduced.

Calibration and Uncertainty Analysis of Sample-Time Error on High Jitter of Samplers

“…The oscilloscope that can measure these high frequency signals is divided into "real-time" and "sampling." The traceable calibration methods for the real-time oscilloscope can be found in [1][2][3][4][5].…”

Characterization of High-Speed Balanced Photodetectors

“…DRTOs contain multiple analog-todigital converters and consequently this creates errors at subNyquist frequencies [13]. The effect of these can be reduced by calibration [14] but these corrections will increase the uncertainties at these frequencies [15]. Alternatively, selecting a comb repetition frequency that fulfills the criteria outlined in [13] will avoid the sub-Nyquist spurs and make the best use of the instrument ADC linearity.…”

Time-interleaved analog-to-digital converter mismatch compensation using infinite impulse response filters

“…Frequency-dependent mismatch correction is important for achieving high-resolution over a wide frequency range and has been studied in various research works . Among these compensation techniques, [104][105][106][107] correct for the bandwidth mismatch whereas [108][109][110][111][112][113][114][115][116][117][118][119][120][121][122][123][124][125][126] deal with the more general frequency response mismatch. One way to avoid performance degradation caused by bandwidth mismatch is to design S/H circuits with bandwidths which are much higher than the maximum input frequency.…”

“…In frequency domain, the sub-ADC outputs and the frequency shifted copies of the input are related by a linear system. Reconstruction in [109] is achieved by computing the inverse of the mismatch matrix.…”

“…The algorithms in [108][109] involve extensive calculations, i.e., DFTs of the sub-ADC outputs and complex matrix inversion, and are therefore not suitable for real time processing. The method presented in [110] compensates for the general frequency response mismatch using the multirate filter bank approach where the signal and aliasing functions on a dense grid of frequency points spanning over the band of interest are controlled simultaneously.…”