Stats

  • Locked Locked
  • Replies 4
  • Subscribers 65
  • Views 3182
  • Members are here 0
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

PSS/Pnoise for Correlated Double Samling

Ashwin Rebello

Hi,

I am analyzing a correlated double sampling system - specific to an image sensor. i am doing 2 conversions and removing the corelated noise in the digital domain by doing a subtraction. In my case the correlate noise comes from a sampled noise at the beginning of the cycle I am using for PSS.

PSS has converged properly, and I have analyzed the time domain signals in the PSS model and they make sense to me.

For Pnoise I have measured the noise (JEE) of the comparator triggering for both conversions and they make sense to me. I used Edge-Crossing for this analysis.

I have then proceeded to measure the time difference between the comparator triggering. using Edge-Delay for Pnoise. I expect the correlated noise which is common to both samples to disappear if I look at the difference. However, that is not the case.

I was wondering if anyone has tried this.

Ashwin

  • There is no pnoise/jitterevent mode to emulate correlated double sampling directly. A possible workaround is to subtract an appropriately delayed copy of the voltage (created by a vcvs with the td parameter set accordingly) from the sampled voltage in the schematic and by performing a pnoise simulation of this difference. However, we found that using the td parameter on the vcvs seems to stop multithreading from working; the additional threads were running idle in this case. We have filed CCR 2625594 for adding correlated double sampling to pnoise; if you ask Cadence Support to file a duplicate on your behalf, this might increase the chances of this being implemented one day.

  • in reply to Frank Wiedmann

    for the pnoise analysis you are referring to above is it the "timeaverage" analysis

  • in reply to Ashwin Rebello

    No, it's the sampled analysis. Here is a simple example netlist where you can see the reduced noise at node rcdiff:

    simulator lang=spectre
    global 0
    R0 (rc 0) resistor r=1M
    C0 (rc 0) capacitor c=1p
    V0 (vsin 0) vsource type=sine freq=1G
    E1 (rcdel rcdiff rc 0) vcvs gain=1.0
    E0 (rcdel 0 rc 0) vcvs gain=1.0 td=500.0p
    simulatorOptions options psfversion="1.4.0" reltol=1e-3 vabstol=1e-6 \
     iabstol=1e-12 temp=27 tnom=27 scalem=1.0 scale=1.0 gmin=1e-12 rforce=1 \
     maxnotes=5 maxwarns=5 digits=5 cols=80 pivrel=1e-3 \
     sensfile="../psf/sens.output" checklimitdest=psf 
    pss pss fund=1G harms=1 errpreset=conservative autotstab=yes
    + annotate=status
    pnoise pnoise start=1 stop=0.5G dec=20 pnoisemethod=fullspectrum \
     maxsideband=1 noisetype=sampled measurement=[pm0 pm1 pm2] \
     annotate=status
     pm0 jitterevent trigger=[vsin 0] triggerthresh=0 triggernum=1 \
      triggerdir=rise target=[rcdiff 0]
     pm1 jitterevent trigger=[vsin 0] triggerthresh=0 triggernum=1 \
      triggerdir=rise target=[rc 0]
     pm2 jitterevent trigger=[vsin 0] triggerthresh=0 triggernum=1 \
      triggerdir=rise target=[rcdel 0]
    saveOptions options save=allpub
  • in reply to Frank Wiedmann

    thanks for your suggest and test case. I implemented the same concept in my test bench. I did a pnoise on the difference between the delayed sample and the current sample. The results make sense to me. 

    I will try and back up your request to add corelated double sampling to pnoise which I think will be useful to many.

Community Guidelines

The Cadence Design Communities support Cadence users and technologists interacting to exchange ideas, news, technical information, and best practices to solve problems and get the most from Cadence technology. The community is open to everyone, and to provide the most value, we require participants to follow our Community Guidelines that facilitate a quality exchange of ideas and information. By accessing, contributing, using or downloading any materials from the site, you agree to be bound by the full Community Guidelines.