ams dynamic para sim question

Yes - see here. Your ADE UI is not the most recent, but it certainly works (and as far as I know has worked for a long time). Why do you think it's not working? Is there an error or a message in the log file? Which versions of Virtuoso, Xcelium and Spectre are you using?

i am doing the ams sim.  so the ADE UI is not same as the spectre

virtuoso 6.1.7

The GUI showed was for AMS - but from IC6.1.8 ISR32. If I remember right, the tran analysis in AMS took longer to change to the newer interface that you see with spectre, but that didn't mean that it wasn't supported.

I would expect it to show up in the log file if it's having an effect. Why do you think this is not supported?

Andrew

why the ams ui is not same with you .

sorry , i test is ok . but why the effective is 1e-05  ?

chase123 said:

why the ams ui is not same with you .

I explained this already. You're using an old, unsupported version (at least 5 years old). The UI has been improved since then.

Of course, I was using AMS as the simulator (strangely enough, I have a pretty reasonable idea what I'm doing...)

chase123 said:

sorry , i test is ok . but why the effective is 1e-05  ?

When you set errpreset=conservative, it sets the reltol to 0.1 of the value you set (if you set it to liberal, it's 10x the value of reltol you set). The log file reports the effective reltol.

Andrew

thank you very much

i want to know

i set conervative +8m 1e-4

  and liberal        +8m 1e-6   

whether the result after 8ms is the same.

then before 8ms it will running fast

thank you very much

i want to know

i set conervative +8m 1e-4

  and liberal        +8m 1e-6   

whether the result after 8ms is the same.

then before 8ms it will running fast

No, it won't be the same because errpreset controls more than just reltol.

Please look at the documentation for more details (or even "spectre -h tran" from the command line for an overview at the end of what errpreset affects).

Andrew

ok.think you very much