Data Weighted Averaging for Simulink
What’s new?
The package still has basic DWA models. Now the models are run by a matlab script.
The stimulus is now created in matlab domain. The truly new thing is the estimation of spurious
tones. This is done by using the stimulus and the DNL error vector shape. See more info at the
last page.
How to simulate
All the parameters and configurations are assigned in run_dwa.m, which also runs the Simulink
model. The Delta-Sigma ADC has fixed oversampling ratio of 32. Use Schreier’s delta-sigma
toolbox (File ID: #19) to alter the parameters in loop filter.
The simulation model DWAMOD_ADC.mdl (Fig. 1) consists of a multibit Delta-Sigma ADC
with DWA mismatch shaping in DAC feedback. The Delta-sigma converter can be either low-
pass- or bandpass-type. There are four models: Two for DWA and two for IDWA, DAC and ADC
for each.
For comparison purposes, we have 2 models for so-called incremental DWA (IDWA). This
means that there are one extra unit element. More about incremental DWA:
K.-D. Chen, and T.-H. Kuo. An Improved Technique for Reducing Baseband Tones in Sigma-
Delta Modulators Employing Data Weighted Averaging Algorithm Without Adding Dither.
IEEE Trans. Circuits and Systems II, Vol. 46, no. 1, pp:63 - 68, Jan. 1999.
Figure 1: A multibit Delta-Sigma ADC. Output “mmnoise” is the mismatch noise.
Compatibility
All models
were saved in Simulink R14 compatible form.
Loop
filter
ADC
DAC with
DWA
double
simout
mmnoise
DS out
stimulus