please help explain this weird result in cadence

[maninnet]

i have a filter, bandpass, center frequency at 100kHZ with gain about 120 V/V (from ac simulation), Vdd=3V.

I run a transient simulation with a sin input of 5mv and 100kHZ,

the first time i run 400us, the result is as expected. (normal.gif)
the second time i run 50us, the result is weird. (werid.gif)

please explain what is the problem?

 

[aryajur]

Did you try changing the time step for the simulation? Maybe that will give the proper result..

 

[maninnet]

i already use "conservative" in cadence, which has the best resolution

 

[aryajur]

Hello Maninnet,
Could you post your BPF circuit, I can try the simulation myself.

 

[maninnet]

don't worry, i figure it out.

 

[aryajur]

What was the error?

 

[maninnet]

i was told that this is due to the high gain (mine is 90dB) of the Opamp, and the transient step in cadence is not small enough to resolve it. I previously had similar situation with an a very high gain opamp when i sim it in transient (open loop, standalone, so i finally had to sim it in close loop to fix the issue), and got a similar explanation for that. But in my current case, i just don't understand
why increase the simulation time would fix the probelm. the two attachment are the simulation record correspondnig to 50u and 500u, and i highlight the crucial data inside, can you intepret why longer simulation time will solve the issue?

 

[aryajur]

Somehow the simulation step might be affecting the simulation. You must try in a separate simulator like HSPICE. Also there is a good book by Ken Kundert SPICE for the IC Designer, I have not read the chapter about Transient Analysis yet, but I am sure the answer can be found there. But I think 1 quick thing to do is to try the circuit in another simulator.