| Author |
 Topic  |
|
|
danmc
USA
2 Posts |
 Posted - Jan 05 2011 : 16:42:22
|
Hello,
I'm trying to extract capacitance between different wires in a coplanar bus. One case I'm looking at is:
Traces are 0.5um wide, 100um long, 16 of them on the same plane with 0.5um spacing between each.
A different case is 25um wide x 1mm long, then a 1um gap, then a 0.5um wide x 1mm long trace, another 1um gap, and a 2nd 25um wide x 1mm long trace. In this case, I'm especially interested in how much capacitance there is between the two outer traces.
I've been trying to use cubegen to generate the traces and then include them in a list file. This seems to work and
fastcap -m -lmyfile.list
produces a graphic that looks right. The part which bothers me is that there seems to be way more discretization along the length of the line. In other words, my 100um of length is chopped up into many (a few hundred?) segments and I thus end up with a very large problem.
Is there some magic flag I need to be feeding cubegen to help with this? X is my width, Y is the length, and Z is perpendicular to the plane the traces reside in.
Many thanks.
-Dan
|
|
|
danmc
USA
2 Posts |
Posted - Jan 05 2011 : 17:41:09
|
| I also forgot to mention that I'm getting a C matrix with some non-negative off diagonal terms. |
 |
|
|
Enrico
550 Posts |
Posted - Jan 08 2011 : 11:46:53
|
Try cubegen_ instead of cubegen.
The difference between the two is exactly the capability to specify the number of panels along each of the three dimensions separately.
(Standard cubegen is maintained for compatibility)
About the non-negative off diagonal, please see the FAQ section for some hints.
Best Regards,
Enrico
|
|
|
| |
Topic |
|
|
|
cubegen issues
