Verification of quantization efficiency and linearity of input-output response in digitizing for radio astronomical observations

An astronomical signal can be addressed as a normal random number that follows a Gaussian distribution of its amplitude. The signal is finally sampled, quantized and digitized before spectroscopic or correlation processing at radio telescope. The quantization efficiency is maximized by adjusting the input power to digitizer. The input power to digitizer has been designed to be within the range where the quantization efficiencies meet the sensitivity requirement in radio astronomical observations. However, we must consider not only the sensitivity maximum but also the accuracy or distortion of waves in digitizing. In this paper, we will report the verification tests about the quantization efficiency and the linearity of the input-output response in digitizing. To meet the high precise observation requirement, the digitizer’s input power should be designed to be much lower than the input power at the highest quantization efficiency by taking into consideration the linearity response.