tran noise sim in cadence of 2-1 sigma-delta modulator with different sampling cap

[xuedashun]

I designed a 2-1 sigma-delta modulator and found fft issue I don't understand. My sim has 2 steps:
1) run "tran" analysis in spectre (select "Transient Noise" option)
2) read data from modulator output to matlab and do fft in matlab

I used 2 different sampling capacitors (1pF, 4pF) for the first integrator and other parameter (fclk, fsample, Nfft=8192, OSR...) are all same.

I use the tran analysis in cadence and select the "Transient Noise" option, the "Noise fmax" is set to 50*fclk. I think this option will take the device noise into account in simulation. But the results are confusing, the 2 different sampling capacitors (1pF and 4pF) do NOT have significant difference on the fft plots and the SNR values only have 2dB difference.

Could anyone give me some suggestions or explain this for me? Thank you!

 

[JoannesPaulus]

In the transient noise form, did you select "Noise update" "step"?

 

[xuedashun]

JoannesPaulus,

Thanks for your reply.

I did not select the "Noise update" "step" in my simulation. Should I select it? What' s the meaning of "Noise update" "step" ???

In the transient noise form, did you select "Noise update" "step"?

 

[JoannesPaulus]

It updates the noise contributions every time the circuit changes operating point (as in a sampling switch).

 

[xuedashun]

JoannesPaulus,

I did not select the "Noise update" "step" option, does that mean the device noise is NOT included in the simulation results so the fft plots of
two cases (Cs=1pF and 4pF) and SNR values are almost same and does not reflect the real noise difference ? Thank you.

It updates the noise contributions every time the circuit changes operating point (as in a sampling switch).

 

[JoannesPaulus]

From Cadence user manual:

Several options are available for updating noise calculation using
noiseupdate parameter :
- noiseupdate=fmax (default)
Noise sources are updated with fixed time interval ΔT = 1/(2*Fmax). Random number and bias dependent amplitude are updated at the same time.
Noise updates will force uniform time step ΔT, so that Tmax= ΔT
- noiseupdate=step
In addition to the regular noise updates at ΔT, amplitude of noise sources is updated at any intermediate time steps.
Time step cannot exceed ΔT, because of regular noise updates (Tmax= ΔT)

So, your noise is included in the FFT plot but it is incorrectly calculated.