Thursday, April 30, 2015

CMRF8SF: bondpad bad device

It took me 3 weeks to completely port a FPGA from cmos14soi to cmrf8sf, so let's say that I generally know what I'm doing in the world of VLSI layout. I have always found that the padframes take the longest of any component to DRC/LVS clean, and this is the single most under-estimated component. Part of this is my fault as I optimize my padframe and pads for different performance because I'm stuck with a few packages. If you are getting strange "bad device" errors in your bondpad cell, make sure that the M1 at the bondpad edge is connected by a contiguous ring of "BFMOAT IND". My corner cells causes a whole padframe to not LVS. 2 days to figure that out.

No comments:

Post a Comment