The complex decimation operation is illustrated with an example in Figure 8-15. First the input signal (and the negative image) are frequency shifted by the NCO frequency as shown on the left. Next a digital filter is applied (centered around 0 Hz) and the output data rate is decimated - in this example the output data rate F_S,OUT = F_S/8 with a Nyquist zone of F_S/16. During the complex mixing the spectrum (signal and noise) is split into real and complex parts and thus the amplitude is reduced by 6-dB. In order to compensate this loss, there is a 6-dB digital gain option in the decimation filter block that can be enabled via SPI write.

Figure 8-15 Complex decimation illustration

The real decimation operation is illustrated with an example in Figure 8-16. There is no frequency shift happening and only the real portion of the complex digital filter is exercised. The output data rate is decimated - a decimation of 8 would result in an output data rate F_S,OUT = F_S/8 with a Nyquist zone of F_S/16.

During the real mixing the spectrum (signal and noise) amplitude is reduced by 3-dB. In order to compensate this loss, there is a 3-dB digital gain option in the decimation filter block that can be enabled via SPI write.

Figure 8-16 Real decimation illustration