The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> Circuit Simulators >> Spectre results become inacurrate for long simulation
https://designers-guide.org/forum/YaBB.pl?num=1435502941

Message started by SimonC on Jun 28^th, 2015, 7:49am

Title: Spectre results become inacurrate for long simulation
Post by SimonC on Jun 28^th, 2015, 7:49am

Hi,
I am simulating a 10 bit ADC with Cadence AMS simulator using Spectre as the solver for the analog part. When I simulate the design for a short period 100 us, the output is as expected. However, when I increase the simulation time above 1 ms, the output becomes inaccurate. It displays some kind of ripple at about twice the sampling frequency. I attached screen shot of the two cases. The purple waveform is when I set the simulator to simulate for 1.024us. The yellow waveform if when I set the simulator for 65 us.
I already tried to switch between pure Spectre and APS mode. I also tried to reduce the max_step parameter down to 100 ps without any impact on the simulator output (except increasing simulation time). I am strobing the output at 1 us interval since it is a MS/s ADC. What parameter/option should I use to tell Spectre to be as accurate for a short and long simulaton?

Title: Re: Spectre results become inacurrate for long simulation
Post by SimonC on Jun 28^th, 2015, 8:13am

I solved the issue. I had and ideal 10 bit DAC verilogA model at the output of the converter. It turns out that the default rise and fall time of the transition filter (transition(out, td, ts, tf)) for the output were set to 0. By setting them to a small value everything is correct now.