A problem about mapping

[Tao Chen]

I did the following things:
1, run "dc" simulations in ADE;
2, get the transconductance of a transistor by this command: OP("/I2/
M37" "gm"), it works well;
3, generate the ocean scripts from ADE;
4, edit the ocean scripts, change the resultsDir from ".../..../
schematic" to ".../.../schematic00", keep all others to be the same,
and run the scripts;
5, select the results "schematic00", run the command OP("/I2/M37"
"gm"), it gives nil
6, run the command OP("I2.M37.mds" "gm"), it gives the results.

My question is: Why I cannot use OP("/I2/M37" "gm")? In the second
simulation, everything in the netlist and map and amap are the same,
spectre should be able to get the mapping information, shouldn't it?

Thanks!

 

[Andrew Beckett]

On Tue, 24 Jul 2007 22:20:16 -0000, Tao Chen <chentao74@gmail.com> wrote:

I did the following things:
1, run "dc" simulations in ADE;
2, get the transconductance of a transistor by this command: OP("/I2/
M37" "gm"), it works well;
3, generate the ocean scripts from ADE;
4, edit the ocean scripts, change the resultsDir from ".../..../
schematic" to ".../.../schematic00", keep all others to be the same,
and run the scripts;
5, select the results "schematic00", run the command OP("/I2/M37"
"gm"), it gives nil
6, run the command OP("I2.M37.mds" "gm"), it gives the results.

My question is: Why I cannot use OP("/I2/M37" "gm")? In the second
simulation, everything in the netlist and map and amap are the same,
spectre should be able to get the mapping information, shouldn't it?

Thanks!

Sounds a bit odd. BTW, it's not spectre that needs the mapping, but OCEAN (or
rather then underlying results access functions). Are you opening the results,
or re-running the simulation?

Probably you should contact customer support over this one - preferrably with
the data you're using to ensure it's well understood.

Regards,

Andrew.
--
Andrew Beckett
Senior Solution Architect
Cadence Design Systems, UK.