|
|
Magic Mailing List |
|
From: cfk (cfk AT pacbell DOT net) Date: Sun Sep 28 2003 - 12:38:58 EDT
I am having trouble understanding painting a genericcontact. Let me tell you where I am and see if you'all can tell me where to go. I paint 2 metal areas, each 8 lambda wide and cross them at 90 degrees with three different technologies. These three are the default scmos, 0.25um and 0.18um described in scmos.tech27, SCN5M_DEEP.12.tech27 and SCN6M_DEEP.09.tech27. I then make a box four lambda wide in the middle of the intersection of metal1 and metal2. Within that box, I then paint genericcontact. I get a different result with different technology files. For instance, if I invoke 'magic -w' to use the default technology and draw my crossing rectangles and the genericcontact, there is no DRC error. If I now do the same thing by invoking with either 'magic -w -TSCN5M_DEEP.12' or 'magic -w -TSCN6M_DEEP.09", I can draw the same rectangles. However, when I draw my 4 x 4 generic contact in the middle, I get an error "Poly, Diffusion overlap of GC contact <1 (Mosis #5.2)" with either 0.25um (SCN5M_DEEP.12) or 0.18um (SCN6M_DEEP.09) technology file. If I now go into scmos.tech27 and search for 'genericcontact', I find an entry in the 'types' section that says: metal1 genericcontact, gcontact, gc If I search into SCN5M_DEEP.12.tech27 I find a somewhat different entry in the 'types' section that says: contact genericcontact, gcontact, gc It gets a bit more interesting in that there is a 'contact' entry in the planes section in SCN5M_DEEP.12.tech27, but no entry for 'contact' in the planes section of the default technology scmos.tech27. So, at this point, what in the woild is going on. Why would the SCN5M_DEEP.12 technology care about poly when I am trying to connect two metal layers together? What am I missing? Is this a magic issue or a Mosis technology file issue? p.s. On Graham's genericcontact issue, just making the contact a little bigger makes the drc's go away. What I dont know is how to do that automatically across all the cells he has created. With Puzzlement (as usual), Charles <sigh>
|
|
|
|