B0008-03
+
Z–1
+
8-Level Quantizer
Z–1
IN
8 fS
OUT
64 fS
+
Z–1
+
Z–1
+
SLES146A – AUGUST 2005 – REVISED OCTOBER 2010
www.ti.com
THEORY OF OPERATION
The DAC section of the PCM1602A is based on a multi-bit delta-sigma architecture. This architecture uses a
fourth-order noise shaper and an 8-level amplitude quantizer, followed by an analog low-pass filter. A block
diagram of the delta-sigma modulator is shown in
Figure 36. This architecture has the advantage of stability and
improved jitter tolerance, when compared to traditional 1-bit (2-level) delta-sigma designs.
Figure 36. Eight-Level Delta-Sigma Modulator
The combined oversampling rate of the digital interpolation filter and the delta-sigma modulator is 32 fS, 64 fS, or
128 fS. The total oversampling rate is determined by the desired sampling frequency. If fS ≤ 96 kHz, then the
OVER bit in register 12 can be set to an oversampling rate of 64 fS or 128 fS. If fS > 96 kHz, then the OVER bit
can be used to set the oversampling rate to 32 fS or 64 fS. Figure 37 shows the out-of-band quantization-noise plots for both the 64× and 128× oversampling scenarios. Notice that the 128× oversampling plot shows
significantly improved out-of-band noise performance, allowing for a simplified low-pass filter to be used at the
output of the DAC.
36
Copyright 2005–2010, Texas Instruments Incorporated