|
|
Magic Mailing List |
|
From: Graham Petley (graham.petley AT vlsitechnology DOT org) Date: Sun Dec 21 2003 - 06:02:49 EST
Hallo,
I am documenting the next release of my Magic library. One
change to the previous release is in the generation of the N
and P-implants. This can be done two ways:
- oversizing and merging of NDIF and PDIF in the different
wells so that implant is a bloated version of diffusion.
- cutting out substrate and well ties from the well layers,
copying to implant and shrinking a little.
The scmos technology file uses the first approach. The
problem with this is that the implant layers do not cover all
of the poly. According to me, for DSM technologies, it is a
requirement that the poly is covered by one or other of the
implants.
My old technology file used the second approach. This gave
implants that covered the poly, but also gave large implants
which extended into adjacent cells. In those cells, the
implants could overlap opposite type implant used for a well
tie, and this is a rule violation. If the implant was shrunk
further to avoid this overlap risk, then the poly sitting at
the nwell/pwell boundary was no longer covered by an implant.
This is shown in the drawing below (monospaced font needed).
nwell----------------------------
pplus.............+
.
4L .
/----/ .
pdif-------+ +-.--ndif
| | .
| | .
| | .
pdif-------+ +-.--ndif
.
.
.
pplus.............+
/------/6L
nwell-----------------------------
The problem is that the diffusion bloat method isn't
adequate when poly must be covered by an implant, and using the
wells isn't adequate because the wells extend into adjacent
cells, which the implant cannot do. I would like to know
whether any magic users out there have a solution to this
problem.
What I have done is include an abutment box layer (AB) into
the layout, and I use this instead of the wells to generate the
implants. I couldn't find an AB layer defined in the existing
Magic technology files, so I added it to my file. How do Magic
users know how close cells can be instantiated if there is no
AB?
Buon natale e felice anno nuovo, Graham
=====
Graham Petley
ASIC Specialist and Consulting Engineer
Metzstrasse 14, 81667 Munich, Germany
Phone +49-89-6349-7630 Mobile +49-177-282-5072
http://www.vlsitechnology.org/
|
|
|
|