dft of sigma delta modulator (cadence analog environment)

[mitrobge]

cadence sigma-delta modulator

I implement a 2nd order Sigma-Delta modulator implemented with verilog-A modules (i.e the modulator is ideal).Plotting the dft of differential output I observe very different results for different time intervals (always taking one period and 256*8 samples).I usually simulate 10 periods (trans analysis).Any ideas about cadence's fourier function accuracy?

 

[field_catcher]

Re: dft of sigma delta modulator (cadence analog environment

For sigma delta modulator, the output is a bit stream containing the quantization noise and the signals ( DC or sinewave ). PSD Estimation is the right way to reveal the information.
Assume your sample frequency is 10M, your run 1.8ms, then you will get 18,000 points, do PSD on last 16,384 points (I forgot if you can do it in Cadence Calculator, but surly you can save the data and export to matlab to do it), you should be able to see 40dB/dec noise shaping if your circuit is constructed correctly.