In other words, there is no downsampling and upsampling. Note that the sample rate remains constant throughout the filter. For more interesting information on complex (quadrature) techniques, I can recommended this blog. After filtering, the complex up-conversion process converts the signal back to the input frequency. The cut-off frequency of both filters is equal to half that of the desired bandpass width. The two lowpass filters are identical, and use the MIFIR (Matrix IFIR) configuration which was described in my previous blog. They are multiplied by the input to give baseband signals to each of the lowpass filters. Sine and cosine waveforms at the filter center frequency have the same sample length as the input signal vector. The input signal, x(n), is processed as a vector of samples (eg from a '.wav' file), which are complex down-converted to baseband by two multipliers. This document gives the formula for the bandpass filter of Figure 1 as follows : See Appendix A for a more detailed view of signal data within the filter.Ī look at the filter block diagram above (Figure 1), which is takenįrom a filter design document by Momentum Data Systems. The filter has complex down-conversion to baseband followed by filtering then up-conversion. This article is available in PDF format for easy printingįig 1.
0 Comments
Leave a Reply. |