- © 2016 by the Seismological Society of America
This article describes the performance tests of three different digitizers that are currently in use by the Idaho National Laboratory seismic network and other seismic networks worldwide. The three digitizers are the PAR4CH from Symmetric Research, and the DAQ24USB‐5V and DAQ24USB‐XR from DAQSystems Inc. The theoretical response of each digitizer is discussed, followed by a description of how the filter coefficients were calculated. The empirically determined sensitivities are compared with the sensitivities reported by the manufacturers. Noise measurements were conducted to better understand the limitations of these digitizers. The digital filters used by all three digitizers are a combination of various sinc N filters; therefore, the filter coefficients were calculated by the convolution of N unit rectangle functions. The sensitivity as a function of the gain was measured and compared with the reported manufacturer’s sensitivities. The DAQ24USB‐5V and DAQ24USB‐XR boards are highly accurate (<0.6%) for gain settings between 1 and 16 but less accurate (>1%) for the higher gain settings of 32 and 64. The PAR4CH board is the least accurate (∼5%) of the three boards in this study. Noise measurements indicate that at lower sampling rates, 100 samples per second, the DAQ24USB‐5V shows the lowest self‐noise and is flat from 1 to 20 Hz, whereas the DAQ24USB‐XR has higher self‐noise at lower frequencies but is flat after 10 Hz. The PAR4CH has the flattest noise model, but the self‐noise is about 30 dB greater than that of the DAQ24USB‐5V. At higher sample rates, around 500 samples per second, the DAQ24USB is very flat up to about 200 Hz, after which it begins to drop, whereas the DAQ24USB‐XR is relatively flat from 10 Hz till Nyquist, thus indicating that the DAQ24USB‐XR is more suitable for applications that require higher sampling rates.