ac and transient simulation in s domain

If you're using the laplace functions in Verilog-A, or using the svcvs (or other similar s-domain controlled sources) in spectre, then transient will work fine. Nothing special you need to do...

Regards,

Andrew.

Unknown said:

Thanks for your reply, but I've tried whatever you told, still it is not working.

"Not working" is rather non-specific. Unfortunately my extra-sensory-perception circuit is broken at the moment so I don't know what your actual problem is. Of course, if you'd like to provide more details, maybe one of us will be able to determine what the problem is?

Regards,

Andrew.

I also see a similar problem, I have a veriloga model using the laplace function which works fine in an ac analysis, however in the transient simulation it causes a blowup failure or with some circuit changes on the testbench I can get it to run but the outputs come back close to zero.

 Inside the veriloga model, I have:

analog begin

@(initial_step("static")) begin

......

nn[0] = (PZ*2*3.14159) ; // These assignment
nn[1] = -1 ; // statements run only
dd[0] = (P1*P2*4*(3.14159**2)) ; // at the beginning of
dd[1] = -((P1+P2)*2*3.14159) ; // the analyses.
dd[2]= 1 ;

gain = (PG*P1*P2*2*3.14159/PZ);

end

V(outp, outn) <+(gain* laplace_nd(V(inp,inn), nn, dd)) ;

end

 In the transient simulation, I see the following:

ERROR (SPECTRE-16384): Signal V(I11:laplace0_state_1) = -1.04665 GV exceeds the blowup limit 

 Also interested in the extra-sensory_perception circuit, is this a Cadence supported feature ;-) ?

Unfortunately I think I'd need to see a complete example (not only the full model, but also how it's being used in the actual circuit where it blows up) to be able to debug. That would be best done via customer support as my guess is that this won't be a 5 minute task...

Unfortunately we don't have the ESP feature yet as standard. It would really make life easier!

Regards,

Andrew.