The transition edge sensor (TES) bias waveform is the summation of carriers of equal, or similar, power. Because of the relative phases, the amplitude of the waveform at any given time sample will exceed the maximum range of the digital-toanalog converter (DAC). And this poses the problem of clipping buffer. Scale the waveform so that it fits within the DAC limits results in carrier and average powers that are reduced by the square of the scaling factor. Directly clip the waveform reduces the carrier power and introduces higher relative phase error. This work develops a clipping algorithm that can generate reliable buffer waveforms without affecting power in each carrier. In TES readout, we need to acquire the accurate initial phases of signals of interest. Coarse channelization followed by fine channelization can be adopted to achieve higher spectral resolution. In radio astronomy, the fast Fourier transform (FFT) is commonly used as channelization for signal processing, while it causes spectral leakage. And this problem can be greatly improved by the technique of polyphase filterbanks (PFB). The PFB consists of a polyphase FIR frontend and an FFT. In order to verify the effectiveness of PFB, we use FFT + FFT method and PFB + FFT method to acquire the initial phases and make a comparison.
|