|
|
Magic Mailing List |
|
From: Jeff W. Sondeen (sondeen AT rcf-fs DOT usc.edu) Date: Mon Aug 06 2001 - 12:20:50 EDT
Siamak Modjtahedi writes: > I should have clarified a point. I knew structures like m123c can be > defined in magic. My question is if things like m1234c can also be > defined? Or in general m[n...n+m]c can be defined? Do you guys know if > this possible? As far as I know the process technology I use allows it but > do not know how to define them in magic. And I tried using m123c and m345c > on top of each other but got a DRC. > magic does have the limitation that vias (actually it's symbolic contacts) can only connect either (only) 2 or 3 adjacent layers, which is why something like m1234c has to be composed out of the stackings of vias that don't erase (replace) intermediate layers, such as m123c + m4c. (if you place m234c on top of m123c, for example, magic erases the m123c since it thinks you changed your mind. i always forget the stackings that work for long stacks and so use macros to build them). you can always forget the symbolic contacts (like m4c), and place so-called 'generic' vias (like gv3), but there is another limitation of magic that resistnace extraction (extres) won't work with generic vias (the 'extres' code uses the actual 'contacts' section of the techfile in which 'generic' vias don't appear). /jeff
|
|
|
|