Stats

  • Locked Locked
  • Replies 1
  • Subscribers 125
  • Views 14570
  • 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

Include CDF callback parameters in netlist

Jan Cools

Dear All,

I'm completing a test design using the gpdk045 and gm/ID methodology for educational purpose. For my technology generation (lookup tables of DC sweeps) I use the gm/ID starter kit scripts of Boris Murmann freely available on the net which I have adapted for the gdpk045. This enables quickly looking at design trade-offs in Matlab. However the results of a simple single transistor simulation between Virtuoso & Matlab are different.

This is because the area & perimeter parameters are not included in my netlist, but are included in a Virtuoso simulation. For a difference in netlist see below:

mn       (vdn vgn 0 vbn) g45n1svt  l=length*1e-6 w=5e-6 nf=5  <-- how it is scripted in Matlab netlist generation

NM3    (net09 net09 0 0) g45n1svt w=(5u) l=45n nf=5 as=460f ad=460f ps=6.92u \   <-- How virtuoso simulation netlist looks like
pd=6.92u nrd=18.4m nrs=18.4m sa=140n sb=140n sd=160n sca=72.26287 \
scb=0.06569 scc=0.00861 m=(1)

These CDF parameters: as, ad, ps, pd, nrd, .. are calculated using callback functions which depend on the transistor dimensions. How can I use these callback functions to extend my Matlab netlist so I can achieve accurate lookup tables?

Best regards

  • Andrew Beckett

    I managed to find the kit (would have made it easier if you'd actually given a link - it's here). The simple answer is no - since it runs completely outside Virtuoso, it wouldn't be able to see the CDF callbacks - for that a component in Virtuoso would need to be created.

    However, looking at what the code is doing, I don't think as, ad, ps and pd matter (since these just affect the capacitance and so don't affect the dc results) - although they are easy enough to compute given that they are just the area and perimeter of the source/drain regions. The stress parameters sa, sb and sd are unaffected by the width and length of the device - so you can find the fixed values from an experiment in Virtuoso and put them in your generated netlist. The well-proximity effect parameters sca, scb and scc are unaffected by the length (which is what the Matlab code seems to sweep) and are fixed for a specific width. From what I can see, this Matlab code generates data for a given width - so presumably you could just see what values are computed by the callbacks for that width and put them into your generated netlist.

    For any more than that, it would be best to contact the author of this "kit".

    Regards,

    Andrew.

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.